_ Git - youtube-dl/blob - youtube_dl/extractor/youtube.py

   1 # coding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5
   6 import itertools
   7 import json
   8 import os.path
   9 import random
  10 import re
  11 import time
  12 import traceback
  13
  14 from .common import InfoExtractor, SearchInfoExtractor
  15 from ..jsinterp import JSInterpreter
  16 from ..swfinterp import SWFInterpreter
  17 from ..compat import (
  18     compat_chr,
  19     compat_HTTPError,
  20     compat_kwargs,
  21     compat_parse_qs,
  22     compat_urllib_parse_unquote,
  23     compat_urllib_parse_unquote_plus,
  24     compat_urllib_parse_urlencode,
  25     compat_urllib_parse_urlparse,
  26     compat_urlparse,
  27     compat_str,
  28 )
  29 from ..utils import (
  30     bool_or_none,
  31     clean_html,
  32     dict_get,
  33     error_to_compat_str,
  34     extract_attributes,
  35     ExtractorError,
  36     float_or_none,
  37     get_element_by_attribute,
  38     get_element_by_id,
  39     int_or_none,
  40     mimetype2ext,
  41     orderedSet,
  42     parse_codecs,
  43     parse_duration,
  44     qualities,
  45     remove_quotes,
  46     remove_start,
  47     smuggle_url,
  48     str_or_none,
  49     str_to_int,
  50     try_get,
  51     unescapeHTML,
  52     unified_strdate,
  53     unsmuggle_url,
  54     uppercase_escape,
  55     url_or_none,
  56     urlencode_postdata,
  57 )
  58
  59
  60 class YoutubeBaseInfoExtractor(InfoExtractor):
  61     """Provide base functions for Youtube extractors"""
  62     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  63     _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
  64
  65     _LOOKUP_URL = 'https://accounts.google.com/_/signin/sl/lookup'
  66     _CHALLENGE_URL = 'https://accounts.google.com/_/signin/sl/challenge'
  67     _TFA_URL = 'https://accounts.google.com/_/signin/challenge?hl=en&TL={0}'
  68
  69     _NETRC_MACHINE = 'youtube'
  70     # If True it will raise an error if no login info is provided
  71     _LOGIN_REQUIRED = False
  72
  73     _PLAYLIST_ID_RE = r'(?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)[0-9A-Za-z-_]{10,}'
  74
  75     def _set_language(self):
  76         self._set_cookie(
  77             '.youtube.com', 'PREF', 'f1=50000000&hl=en',
  78             # YouTube sets the expire time to about two months
  79             expire_time=time.time() + 2 * 30 * 24 * 3600)
  80
  81     def _ids_to_results(self, ids):
  82         return [
  83             self.url_result(vid_id, 'Youtube', video_id=vid_id)
  84             for vid_id in ids]
  85
  86     def _login(self):
  87         """
  88         Attempt to log in to YouTube.
  89         True is returned if successful or skipped.
  90         False is returned if login failed.
  91
  92         If _LOGIN_REQUIRED is set and no authentication was provided, an error is raised.
  93         """
  94         username, password = self._get_login_info()
  95         # No authentication to be performed
  96         if username is None:
  97             if self._LOGIN_REQUIRED and self._downloader.params.get('cookiefile') is None:
  98                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  99             return True
 100
 101         login_page = self._download_webpage(
 102             self._LOGIN_URL, None,
 103             note='Downloading login page',
 104             errnote='unable to fetch login page', fatal=False)
 105         if login_page is False:
 106             return
 107
 108         login_form = self._hidden_inputs(login_page)
 109
 110         def req(url, f_req, note, errnote):
 111             data = login_form.copy()
 112             data.update({
 113                 'pstMsg': 1,
 114                 'checkConnection': 'youtube',
 115                 'checkedDomains': 'youtube',
 116                 'hl': 'en',
 117                 'deviceinfo': '[null,null,null,[],null,"US",null,null,[],"GlifWebSignIn",null,[null,null,[]]]',
 118                 'f.req': json.dumps(f_req),
 119                 'flowName': 'GlifWebSignIn',
 120                 'flowEntry': 'ServiceLogin',
 121                 # TODO: reverse actual botguard identifier generation algo
 122                 'bgRequest': '["identifier",""]',
 123             })
 124             return self._download_json(
 125                 url, None, note=note, errnote=errnote,
 126                 transform_source=lambda s: re.sub(r'^[^[]*', '', s),
 127                 fatal=False,
 128                 data=urlencode_postdata(data), headers={
 129                     'Content-Type': 'application/x-www-form-urlencoded;charset=utf-8',
 130                     'Google-Accounts-XSRF': 1,
 131                 })
 132
 133         def warn(message):
 134             self._downloader.report_warning(message)
 135
 136         lookup_req = [
 137             username,
 138             None, [], None, 'US', None, None, 2, False, True,
 139             [
 140                 None, None,
 141                 [2, 1, None, 1,
 142                  'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn',
 143                  None, [], 4],
 144                 1, [None, None, []], None, None, None, True
 145             ],
 146             username,
 147         ]
 148
 149         lookup_results = req(
 150             self._LOOKUP_URL, lookup_req,
 151             'Looking up account info', 'Unable to look up account info')
 152
 153         if lookup_results is False:
 154             return False
 155
 156         user_hash = try_get(lookup_results, lambda x: x[0][2], compat_str)
 157         if not user_hash:
 158             warn('Unable to extract user hash')
 159             return False
 160
 161         challenge_req = [
 162             user_hash,
 163             None, 1, None, [1, None, None, None, [password, None, True]],
 164             [
 165                 None, None, [2, 1, None, 1, 'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn', None, [], 4],
 166                 1, [None, None, []], None, None, None, True
 167             ]]
 168
 169         challenge_results = req(
 170             self._CHALLENGE_URL, challenge_req,
 171             'Logging in', 'Unable to log in')
 172
 173         if challenge_results is False:
 174             return
 175
 176         login_res = try_get(challenge_results, lambda x: x[0][5], list)
 177         if login_res:
 178             login_msg = try_get(login_res, lambda x: x[5], compat_str)
 179             warn(
 180                 'Unable to login: %s' % 'Invalid password'
 181                 if login_msg == 'INCORRECT_ANSWER_ENTERED' else login_msg)
 182             return False
 183
 184         res = try_get(challenge_results, lambda x: x[0][-1], list)
 185         if not res:
 186             warn('Unable to extract result entry')
 187             return False
 188
 189         login_challenge = try_get(res, lambda x: x[0][0], list)
 190         if login_challenge:
 191             challenge_str = try_get(login_challenge, lambda x: x[2], compat_str)
 192             if challenge_str == 'TWO_STEP_VERIFICATION':
 193                 # SEND_SUCCESS - TFA code has been successfully sent to phone
 194                 # QUOTA_EXCEEDED - reached the limit of TFA codes
 195                 status = try_get(login_challenge, lambda x: x[5], compat_str)
 196                 if status == 'QUOTA_EXCEEDED':
 197                     warn('Exceeded the limit of TFA codes, try later')
 198                     return False
 199
 200                 tl = try_get(challenge_results, lambda x: x[1][2], compat_str)
 201                 if not tl:
 202                     warn('Unable to extract TL')
 203                     return False
 204
 205                 tfa_code = self._get_tfa_info('2-step verification code')
 206
 207                 if not tfa_code:
 208                     warn(
 209                         'Two-factor authentication required. Provide it either interactively or with --twofactor <code>'
 210                         '(Note that only TOTP (Google Authenticator App) codes work at this time.)')
 211                     return False
 212
 213                 tfa_code = remove_start(tfa_code, 'G-')
 214
 215                 tfa_req = [
 216                     user_hash, None, 2, None,
 217                     [
 218                         9, None, None, None, None, None, None, None,
 219                         [None, tfa_code, True, 2]
 220                     ]]
 221
 222                 tfa_results = req(
 223                     self._TFA_URL.format(tl), tfa_req,
 224                     'Submitting TFA code', 'Unable to submit TFA code')
 225
 226                 if tfa_results is False:
 227                     return False
 228
 229                 tfa_res = try_get(tfa_results, lambda x: x[0][5], list)
 230                 if tfa_res:
 231                     tfa_msg = try_get(tfa_res, lambda x: x[5], compat_str)
 232                     warn(
 233                         'Unable to finish TFA: %s' % 'Invalid TFA code'
 234                         if tfa_msg == 'INCORRECT_ANSWER_ENTERED' else tfa_msg)
 235                     return False
 236
 237                 check_cookie_url = try_get(
 238                     tfa_results, lambda x: x[0][-1][2], compat_str)
 239             else:
 240                 CHALLENGES = {
 241                     'LOGIN_CHALLENGE': "This device isn't recognized. For your security, Google wants to make sure it's really you.",
 242                     'USERNAME_RECOVERY': 'Please provide additional information to aid in the recovery process.',
 243                     'REAUTH': "There is something unusual about your activity. For your security, Google wants to make sure it's really you.",
 244                 }
 245                 challenge = CHALLENGES.get(
 246                     challenge_str,
 247                     '%s returned error %s.' % (self.IE_NAME, challenge_str))
 248                 warn('%s\nGo to https://accounts.google.com/, login and solve a challenge.' % challenge)
 249                 return False
 250         else:
 251             check_cookie_url = try_get(res, lambda x: x[2], compat_str)
 252
 253         if not check_cookie_url:
 254             warn('Unable to extract CheckCookie URL')
 255             return False
 256
 257         check_cookie_results = self._download_webpage(
 258             check_cookie_url, None, 'Checking cookie', fatal=False)
 259
 260         if check_cookie_results is False:
 261             return False
 262
 263         if 'https://myaccount.google.com/' not in check_cookie_results:
 264             warn('Unable to log in')
 265             return False
 266
 267         return True
 268
 269     def _download_webpage_handle(self, *args, **kwargs):
 270         query = kwargs.get('query', {}).copy()
 271         query['disable_polymer'] = 'true'
 272         kwargs['query'] = query
 273         return super(YoutubeBaseInfoExtractor, self)._download_webpage_handle(
 274             *args, **compat_kwargs(kwargs))
 275
 276     def _real_initialize(self):
 277         if self._downloader is None:
 278             return
 279         self._set_language()
 280         if not self._login():
 281             return
 282
 283
 284 class YoutubeEntryListBaseInfoExtractor(YoutubeBaseInfoExtractor):
 285     # Extract entries from page with "Load more" button
 286     def _entries(self, page, playlist_id):
 287         more_widget_html = content_html = page
 288         for page_num in itertools.count(1):
 289             for entry in self._process_page(content_html):
 290                 yield entry
 291
 292             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
 293             if not mobj:
 294                 break
 295
 296             count = 0
 297             retries = 3
 298             while count <= retries:
 299                 try:
 300                     # Downloading page may result in intermittent 5xx HTTP error
 301                     # that is usually worked around with a retry
 302                     more = self._download_json(
 303                         'https://youtube.com/%s' % mobj.group('more'), playlist_id,
 304                         'Downloading page #%s%s'
 305                         % (page_num, ' (retry #%d)' % count if count else ''),
 306                         transform_source=uppercase_escape)
 307                     break
 308                 except ExtractorError as e:
 309                     if isinstance(e.cause, compat_HTTPError) and e.cause.code in (500, 503):
 310                         count += 1
 311                         if count <= retries:
 312                             continue
 313                     raise
 314
 315             content_html = more['content_html']
 316             if not content_html.strip():
 317                 # Some webpages show a "Load more" button but they don't
 318                 # have more videos
 319                 break
 320             more_widget_html = more['load_more_widget_html']
 321
 322
 323 class YoutubePlaylistBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
 324     def _process_page(self, content):
 325         for video_id, video_title in self.extract_videos_from_page(content):
 326             yield self.url_result(video_id, 'Youtube', video_id, video_title)
 327
 328     def extract_videos_from_page_impl(self, video_re, page, ids_in_page, titles_in_page):
 329         for mobj in re.finditer(video_re, page):
 330             # The link with index 0 is not the first video of the playlist (not sure if still actual)
 331             if 'index' in mobj.groupdict() and mobj.group('id') == '0':
 332                 continue
 333             video_id = mobj.group('id')
 334             video_title = unescapeHTML(
 335                 mobj.group('title')) if 'title' in mobj.groupdict() else None
 336             if video_title:
 337                 video_title = video_title.strip()
 338             if video_title == '► Play all':
 339                 video_title = None
 340             try:
 341                 idx = ids_in_page.index(video_id)
 342                 if video_title and not titles_in_page[idx]:
 343                     titles_in_page[idx] = video_title
 344             except ValueError:
 345                 ids_in_page.append(video_id)
 346                 titles_in_page.append(video_title)
 347
 348     def extract_videos_from_page(self, page):
 349         ids_in_page = []
 350         titles_in_page = []
 351         self.extract_videos_from_page_impl(
 352             self._VIDEO_RE, page, ids_in_page, titles_in_page)
 353         return zip(ids_in_page, titles_in_page)
 354
 355
 356 class YoutubePlaylistsBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
 357     def _process_page(self, content):
 358         for playlist_id in orderedSet(re.findall(
 359                 r'<h3[^>]+class="[^"]*yt-lockup-title[^"]*"[^>]*><a[^>]+href="/?playlist\?list=([0-9A-Za-z-_]{10,})"',
 360                 content)):
 361             yield self.url_result(
 362                 'https://www.youtube.com/playlist?list=%s' % playlist_id, 'YoutubePlaylist')
 363
 364     def _real_extract(self, url):
 365         playlist_id = self._match_id(url)
 366         webpage = self._download_webpage(url, playlist_id)
 367         title = self._og_search_title(webpage, fatal=False)
 368         return self.playlist_result(self._entries(webpage, playlist_id), playlist_id, title)
 369
 370
 371 class YoutubeIE(YoutubeBaseInfoExtractor):
 372     IE_DESC = 'YouTube.com'
 373     _VALID_URL = r"""(?x)^
 374                      (
 375                          (?:https?://|//)                                    # http(s):// or protocol-independent URL
 376                          (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/|
 377                             (?:www\.)?deturl\.com/www\.youtube\.com/|
 378                             (?:www\.)?pwnyoutube\.com/|
 379                             (?:www\.)?hooktube\.com/|
 380                             (?:www\.)?yourepeat\.com/|
 381                             tube\.majestyc\.net/|
 382                             # Invidious instances taken from https://github.com/omarroth/invidious/wiki/Invidious-Instances
 383                             (?:(?:www|dev)\.)?invidio\.us/|
 384                             (?:(?:www|no)\.)?invidiou\.sh/|
 385                             (?:(?:www|fi|de)\.)?invidious\.snopyta\.org/|
 386                             (?:www\.)?invidious\.kabi\.tk/|
 387                             (?:www\.)?invidious\.enkirton\.net/|
 388                             (?:www\.)?invidious\.13ad\.de/|
 389                             (?:www\.)?invidious\.mastodon\.host/|
 390                             (?:www\.)?tube\.poal\.co/|
 391                             (?:www\.)?vid\.wxzm\.sx/|
 392                             youtube\.googleapis\.com/)                        # the various hostnames, with wildcard subdomains
 393                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 394                          (?:                                                  # the various things that can precede the ID:
 395                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
 396                              |(?:                                             # or the v= param in all its forms
 397                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 398                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 399                                  (?:.*?[&;])??                                # any other preceding param (like /?s=tuff&v=xxxx or ?s=tuff&amp;v=V36LpHqtcDY)
 400                                  v=
 401                              )
 402                          ))
 403                          |(?:
 404                             youtu\.be|                                        # just youtu.be/xxxx
 405                             vid\.plus|                                        # or vid.plus/xxxx
 406                             zwearz\.com/watch|                                # or zwearz.com/watch/xxxx
 407                          )/
 408                          |(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
 409                          )
 410                      )?                                                       # all until now is optional -> you can pass the naked ID
 411                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
 412                      (?!.*?\blist=
 413                         (?:
 414                             %(playlist_id)s|                                  # combined list/video URLs are handled by the playlist IE
 415                             WL                                                # WL are handled by the watch later IE
 416                         )
 417                      )
 418                      (?(1).+)?                                                # if we found the ID, everything can follow
 419                      $""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
 420     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 421     _formats = {
 422         '5': {'ext': 'flv', 'width': 400, 'height': 240, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
 423         '6': {'ext': 'flv', 'width': 450, 'height': 270, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
 424         '13': {'ext': '3gp', 'acodec': 'aac', 'vcodec': 'mp4v'},
 425         '17': {'ext': '3gp', 'width': 176, 'height': 144, 'acodec': 'aac', 'abr': 24, 'vcodec': 'mp4v'},
 426         '18': {'ext': 'mp4', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 96, 'vcodec': 'h264'},
 427         '22': {'ext': 'mp4', 'width': 1280, 'height': 720, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 428         '34': {'ext': 'flv', 'width': 640, 'height': 360, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 429         '35': {'ext': 'flv', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 430         # itag 36 videos are either 320x180 (BaW_jenozKc) or 320x240 (__2ABJjxzNo), abr varies as well
 431         '36': {'ext': '3gp', 'width': 320, 'acodec': 'aac', 'vcodec': 'mp4v'},
 432         '37': {'ext': 'mp4', 'width': 1920, 'height': 1080, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 433         '38': {'ext': 'mp4', 'width': 4096, 'height': 3072, 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264'},
 434         '43': {'ext': 'webm', 'width': 640, 'height': 360, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'},
 435         '44': {'ext': 'webm', 'width': 854, 'height': 480, 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8'},
 436         '45': {'ext': 'webm', 'width': 1280, 'height': 720, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'},
 437         '46': {'ext': 'webm', 'width': 1920, 'height': 1080, 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8'},
 438         '59': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 439         '78': {'ext': 'mp4', 'width': 854, 'height': 480, 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264'},
 440
 441
 442         # 3D videos
 443         '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20},
 444         '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -20},
 445         '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20},
 446         '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'acodec': 'aac', 'abr': 192, 'vcodec': 'h264', 'preference': -20},
 447         '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 128, 'vcodec': 'vp8', 'preference': -20},
 448         '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20},
 449         '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'acodec': 'vorbis', 'abr': 192, 'vcodec': 'vp8', 'preference': -20},
 450
 451         # Apple HTTP Live Streaming
 452         '91': {'ext': 'mp4', 'height': 144, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 453         '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 454         '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10},
 455         '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 128, 'vcodec': 'h264', 'preference': -10},
 456         '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10},
 457         '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 256, 'vcodec': 'h264', 'preference': -10},
 458         '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 48, 'vcodec': 'h264', 'preference': -10},
 459         '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'acodec': 'aac', 'abr': 24, 'vcodec': 'h264', 'preference': -10},
 460
 461         # DASH mp4 video
 462         '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'h264'},
 463         '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'h264'},
 464         '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'},
 465         '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264'},
 466         '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264'},
 467         '138': {'ext': 'mp4', 'format_note': 'DASH video', 'vcodec': 'h264'},  # Height can vary (https://github.com/ytdl-org/youtube-dl/issues/4559)
 468         '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'vcodec': 'h264'},
 469         '212': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'h264'},
 470         '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'h264'},
 471         '298': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60},
 472         '299': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'h264', 'fps': 60},
 473         '266': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'h264'},
 474
 475         # Dash mp4 audio
 476         '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 48, 'container': 'm4a_dash'},
 477         '140': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 128, 'container': 'm4a_dash'},
 478         '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'abr': 256, 'container': 'm4a_dash'},
 479         '256': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'},
 480         '258': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'aac', 'container': 'm4a_dash'},
 481         '325': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'dtse', 'container': 'm4a_dash'},
 482         '328': {'ext': 'm4a', 'format_note': 'DASH audio', 'acodec': 'ec-3', 'container': 'm4a_dash'},
 483
 484         # Dash webm
 485         '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 486         '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 487         '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 488         '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 489         '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 490         '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp8'},
 491         '278': {'ext': 'webm', 'height': 144, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'vp9'},
 492         '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 493         '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 494         '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 495         '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 496         '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 497         '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 498         '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 499         '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 500         # itag 272 videos are either 3840x2160 (e.g. RtoitU2A-3E) or 7680x4320 (sLprVF6d7Ug)
 501         '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 502         '302': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 503         '303': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 504         '308': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 505         '313': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9'},
 506         '315': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'vcodec': 'vp9', 'fps': 60},
 507
 508         # Dash webm audio
 509         '171': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 128},
 510         '172': {'ext': 'webm', 'acodec': 'vorbis', 'format_note': 'DASH audio', 'abr': 256},
 511
 512         # Dash webm audio with opus inside
 513         '249': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 50},
 514         '250': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 70},
 515         '251': {'ext': 'webm', 'format_note': 'DASH audio', 'acodec': 'opus', 'abr': 160},
 516
 517         # RTMP (unnamed)
 518         '_rtmp': {'protocol': 'rtmp'},
 519
 520         # av01 video only formats sometimes served with "unknown" codecs
 521         '394': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 522         '395': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 523         '396': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 524         '397': {'acodec': 'none', 'vcodec': 'av01.0.05M.08'},
 525     }
 526     _SUBTITLE_FORMATS = ('srv1', 'srv2', 'srv3', 'ttml', 'vtt')
 527
 528     _GEO_BYPASS = False
 529
 530     IE_NAME = 'youtube'
 531     _TESTS = [
 532         {
 533             'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&t=1s&end=9',
 534             'info_dict': {
 535                 'id': 'BaW_jenozKc',
 536                 'ext': 'mp4',
 537                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 538                 'uploader': 'Philipp Hagemeister',
 539                 'uploader_id': 'phihag',
 540                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag',
 541                 'channel_id': 'UCLqxVugv74EIW3VWh2NOa3Q',
 542                 'channel_url': r're:https?://(?:www\.)?youtube\.com/channel/UCLqxVugv74EIW3VWh2NOa3Q',
 543                 'upload_date': '20121002',
 544                 'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
 545                 'categories': ['Science & Technology'],
 546                 'tags': ['youtube-dl'],
 547                 'duration': 10,
 548                 'view_count': int,
 549                 'like_count': int,
 550                 'dislike_count': int,
 551                 'start_time': 1,
 552                 'end_time': 9,
 553             }
 554         },
 555         {
 556             'url': 'https://www.youtube.com/watch?v=UxxajLWwzqY',
 557             'note': 'Test generic use_cipher_signature video (#897)',
 558             'info_dict': {
 559                 'id': 'UxxajLWwzqY',
 560                 'ext': 'mp4',
 561                 'upload_date': '20120506',
 562                 'title': 'Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]',
 563                 'alt_title': 'I Love It (feat. Charli XCX)',
 564                 'description': 'md5:f3ceb5ef83a08d95b9d146f973157cc8',
 565                 'tags': ['Icona Pop i love it', 'sweden', 'pop music', 'big beat records', 'big beat', 'charli',
 566                          'xcx', 'charli xcx', 'girls', 'hbo', 'i love it', "i don't care", 'icona', 'pop',
 567                          'iconic ep', 'iconic', 'love', 'it'],
 568                 'duration': 180,
 569                 'uploader': 'Icona Pop',
 570                 'uploader_id': 'IconaPop',
 571                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IconaPop',
 572                 'creator': 'Icona Pop',
 573                 'track': 'I Love It (feat. Charli XCX)',
 574                 'artist': 'Icona Pop',
 575             }
 576         },
 577         {
 578             'url': 'https://www.youtube.com/watch?v=07FYdnEawAQ',
 579             'note': 'Test VEVO video with age protection (#956)',
 580             'info_dict': {
 581                 'id': '07FYdnEawAQ',
 582                 'ext': 'mp4',
 583                 'upload_date': '20130703',
 584                 'title': 'Justin Timberlake - Tunnel Vision (Official Music Video) (Explicit)',
 585                 'alt_title': 'Tunnel Vision',
 586                 'description': 'md5:07dab3356cde4199048e4c7cd93471e1',
 587                 'duration': 419,
 588                 'uploader': 'justintimberlakeVEVO',
 589                 'uploader_id': 'justintimberlakeVEVO',
 590                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/justintimberlakeVEVO',
 591                 'creator': 'Justin Timberlake',
 592                 'track': 'Tunnel Vision',
 593                 'artist': 'Justin Timberlake',
 594                 'age_limit': 18,
 595             }
 596         },
 597         {
 598             'url': '//www.YouTube.com/watch?v=yZIXLfi8CZQ',
 599             'note': 'Embed-only video (#1746)',
 600             'info_dict': {
 601                 'id': 'yZIXLfi8CZQ',
 602                 'ext': 'mp4',
 603                 'upload_date': '20120608',
 604                 'title': 'Principal Sexually Assaults A Teacher - Episode 117 - 8th June 2012',
 605                 'description': 'md5:09b78bd971f1e3e289601dfba15ca4f7',
 606                 'uploader': 'SET India',
 607                 'uploader_id': 'setindia',
 608                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/setindia',
 609                 'age_limit': 18,
 610             }
 611         },
 612         {
 613             'url': 'https://www.youtube.com/watch?v=BaW_jenozKc&v=UxxajLWwzqY',
 614             'note': 'Use the first video ID in the URL',
 615             'info_dict': {
 616                 'id': 'BaW_jenozKc',
 617                 'ext': 'mp4',
 618                 'title': 'youtube-dl test video "\'/\\ä↭𝕐',
 619                 'uploader': 'Philipp Hagemeister',
 620                 'uploader_id': 'phihag',
 621                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/phihag',
 622                 'upload_date': '20121002',
 623                 'description': 'test chars:  "\'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .',
 624                 'categories': ['Science & Technology'],
 625                 'tags': ['youtube-dl'],
 626                 'duration': 10,
 627                 'view_count': int,
 628                 'like_count': int,
 629                 'dislike_count': int,
 630             },
 631             'params': {
 632                 'skip_download': True,
 633             },
 634         },
 635         {
 636             'url': 'https://www.youtube.com/watch?v=a9LDPn-MO4I',
 637             'note': '256k DASH audio (format 141) via DASH manifest',
 638             'info_dict': {
 639                 'id': 'a9LDPn-MO4I',
 640                 'ext': 'm4a',
 641                 'upload_date': '20121002',
 642                 'uploader_id': '8KVIDEO',
 643                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/8KVIDEO',
 644                 'description': '',
 645                 'uploader': '8KVIDEO',
 646                 'title': 'UHDTV TEST 8K VIDEO.mp4'
 647             },
 648             'params': {
 649                 'youtube_include_dash_manifest': True,
 650                 'format': '141',
 651             },
 652             'skip': 'format 141 not served anymore',
 653         },
 654         # DASH manifest with encrypted signature
 655         {
 656             'url': 'https://www.youtube.com/watch?v=IB3lcPjvWLA',
 657             'info_dict': {
 658                 'id': 'IB3lcPjvWLA',
 659                 'ext': 'm4a',
 660                 'title': 'Afrojack, Spree Wilson - The Spark (Official Music Video) ft. Spree Wilson',
 661                 'description': 'md5:8f5e2b82460520b619ccac1f509d43bf',
 662                 'duration': 244,
 663                 'uploader': 'AfrojackVEVO',
 664                 'uploader_id': 'AfrojackVEVO',
 665                 'upload_date': '20131011',
 666             },
 667             'params': {
 668                 'youtube_include_dash_manifest': True,
 669                 'format': '141/bestaudio[ext=m4a]',
 670             },
 671         },
 672         # JS player signature function name containing $
 673         {
 674             'url': 'https://www.youtube.com/watch?v=nfWlot6h_JM',
 675             'info_dict': {
 676                 'id': 'nfWlot6h_JM',
 677                 'ext': 'm4a',
 678                 'title': 'Taylor Swift - Shake It Off',
 679                 'description': 'md5:bec2185232c05479482cb5a9b82719bf',
 680                 'duration': 242,
 681                 'uploader': 'TaylorSwiftVEVO',
 682                 'uploader_id': 'TaylorSwiftVEVO',
 683                 'upload_date': '20140818',
 684                 'creator': 'Taylor Swift',
 685             },
 686             'params': {
 687                 'youtube_include_dash_manifest': True,
 688                 'format': '141/bestaudio[ext=m4a]',
 689             },
 690         },
 691         # Controversy video
 692         {
 693             'url': 'https://www.youtube.com/watch?v=T4XJQO3qol8',
 694             'info_dict': {
 695                 'id': 'T4XJQO3qol8',
 696                 'ext': 'mp4',
 697                 'duration': 219,
 698                 'upload_date': '20100909',
 699                 'uploader': 'Amazing Atheist',
 700                 'uploader_id': 'TheAmazingAtheist',
 701                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheAmazingAtheist',
 702                 'title': 'Burning Everyone\'s Koran',
 703                 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms\n\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html',
 704             }
 705         },
 706         # Normal age-gate video (No vevo, embed allowed)
 707         {
 708             'url': 'https://youtube.com/watch?v=HtVdAasjOgU',
 709             'info_dict': {
 710                 'id': 'HtVdAasjOgU',
 711                 'ext': 'mp4',
 712                 'title': 'The Witcher 3: Wild Hunt - The Sword Of Destiny Trailer',
 713                 'description': r're:(?s).{100,}About the Game\n.*?The Witcher 3: Wild Hunt.{100,}',
 714                 'duration': 142,
 715                 'uploader': 'The Witcher',
 716                 'uploader_id': 'WitcherGame',
 717                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/WitcherGame',
 718                 'upload_date': '20140605',
 719                 'age_limit': 18,
 720             },
 721         },
 722         # Age-gate video with encrypted signature
 723         {
 724             'url': 'https://www.youtube.com/watch?v=6kLq3WMV1nU',
 725             'info_dict': {
 726                 'id': '6kLq3WMV1nU',
 727                 'ext': 'mp4',
 728                 'title': 'Dedication To My Ex (Miss That) (Lyric Video)',
 729                 'description': 'md5:33765bb339e1b47e7e72b5490139bb41',
 730                 'duration': 246,
 731                 'uploader': 'LloydVEVO',
 732                 'uploader_id': 'LloydVEVO',
 733                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/LloydVEVO',
 734                 'upload_date': '20110629',
 735                 'age_limit': 18,
 736             },
 737         },
 738         # video_info is None (https://github.com/ytdl-org/youtube-dl/issues/4421)
 739         # YouTube Red ad is not captured for creator
 740         {
 741             'url': '__2ABJjxzNo',
 742             'info_dict': {
 743                 'id': '__2ABJjxzNo',
 744                 'ext': 'mp4',
 745                 'duration': 266,
 746                 'upload_date': '20100430',
 747                 'uploader_id': 'deadmau5',
 748                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/deadmau5',
 749                 'creator': 'deadmau5',
 750                 'description': 'md5:12c56784b8032162bb936a5f76d55360',
 751                 'uploader': 'deadmau5',
 752                 'title': 'Deadmau5 - Some Chords (HD)',
 753                 'alt_title': 'Some Chords',
 754             },
 755             'expected_warnings': [
 756                 'DASH manifest missing',
 757             ]
 758         },
 759         # Olympics (https://github.com/ytdl-org/youtube-dl/issues/4431)
 760         {
 761             'url': 'lqQg6PlCWgI',
 762             'info_dict': {
 763                 'id': 'lqQg6PlCWgI',
 764                 'ext': 'mp4',
 765                 'duration': 6085,
 766                 'upload_date': '20150827',
 767                 'uploader_id': 'olympic',
 768                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/olympic',
 769                 'description': 'HO09  - Women -  GER-AUS - Hockey - 31 July 2012 - London 2012 Olympic Games',
 770                 'uploader': 'Olympic',
 771                 'title': 'Hockey - Women -  GER-AUS - London 2012 Olympic Games',
 772             },
 773             'params': {
 774                 'skip_download': 'requires avconv',
 775             }
 776         },
 777         # Non-square pixels
 778         {
 779             'url': 'https://www.youtube.com/watch?v=_b-2C3KPAM0',
 780             'info_dict': {
 781                 'id': '_b-2C3KPAM0',
 782                 'ext': 'mp4',
 783                 'stretched_ratio': 16 / 9.,
 784                 'duration': 85,
 785                 'upload_date': '20110310',
 786                 'uploader_id': 'AllenMeow',
 787                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/AllenMeow',
 788                 'description': 'made by Wacom from Korea | 字幕&加油添醋 by TY\'s Allen | 感謝heylisa00cavey1001同學熱情提供梗及翻譯',
 789                 'uploader': '孫ᄋᄅ',
 790                 'title': '[A-made] 變態妍字幕版 太妍 我就是這樣的人',
 791             },
 792         },
 793         # url_encoded_fmt_stream_map is empty string
 794         {
 795             'url': 'qEJwOuvDf7I',
 796             'info_dict': {
 797                 'id': 'qEJwOuvDf7I',
 798                 'ext': 'webm',
 799                 'title': 'Обсуждение судебной практики по выборам 14 сентября 2014 года в Санкт-Петербурге',
 800                 'description': '',
 801                 'upload_date': '20150404',
 802                 'uploader_id': 'spbelect',
 803                 'uploader': 'Наблюдатели Петербурга',
 804             },
 805             'params': {
 806                 'skip_download': 'requires avconv',
 807             },
 808             'skip': 'This live event has ended.',
 809         },
 810         # Extraction from multiple DASH manifests (https://github.com/ytdl-org/youtube-dl/pull/6097)
 811         {
 812             'url': 'https://www.youtube.com/watch?v=FIl7x6_3R5Y',
 813             'info_dict': {
 814                 'id': 'FIl7x6_3R5Y',
 815                 'ext': 'webm',
 816                 'title': 'md5:7b81415841e02ecd4313668cde88737a',
 817                 'description': 'md5:116377fd2963b81ec4ce64b542173306',
 818                 'duration': 220,
 819                 'upload_date': '20150625',
 820                 'uploader_id': 'dorappi2000',
 821                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/dorappi2000',
 822                 'uploader': 'dorappi2000',
 823                 'formats': 'mincount:31',
 824             },
 825             'skip': 'not actual anymore',
 826         },
 827         # DASH manifest with segment_list
 828         {
 829             'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
 830             'md5': '8ce563a1d667b599d21064e982ab9e31',
 831             'info_dict': {
 832                 'id': 'CsmdDsKjzN8',
 833                 'ext': 'mp4',
 834                 'upload_date': '20150501',  # According to '<meta itemprop="datePublished"', but in other places it's 20150510
 835                 'uploader': 'Airtek',
 836                 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
 837                 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
 838                 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
 839             },
 840             'params': {
 841                 'youtube_include_dash_manifest': True,
 842                 'format': '135',  # bestvideo
 843             },
 844             'skip': 'This live event has ended.',
 845         },
 846         {
 847             # Multifeed videos (multiple cameras), URL is for Main Camera
 848             'url': 'https://www.youtube.com/watch?v=jqWvoWXjCVs',
 849             'info_dict': {
 850                 'id': 'jqWvoWXjCVs',
 851                 'title': 'teamPGP: Rocket League Noob Stream',
 852                 'description': 'md5:dc7872fb300e143831327f1bae3af010',
 853             },
 854             'playlist': [{
 855                 'info_dict': {
 856                     'id': 'jqWvoWXjCVs',
 857                     'ext': 'mp4',
 858                     'title': 'teamPGP: Rocket League Noob Stream (Main Camera)',
 859                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 860                     'duration': 7335,
 861                     'upload_date': '20150721',
 862                     'uploader': 'Beer Games Beer',
 863                     'uploader_id': 'beergamesbeer',
 864                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 865                     'license': 'Standard YouTube License',
 866                 },
 867             }, {
 868                 'info_dict': {
 869                     'id': '6h8e8xoXJzg',
 870                     'ext': 'mp4',
 871                     'title': 'teamPGP: Rocket League Noob Stream (kreestuh)',
 872                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 873                     'duration': 7337,
 874                     'upload_date': '20150721',
 875                     'uploader': 'Beer Games Beer',
 876                     'uploader_id': 'beergamesbeer',
 877                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 878                     'license': 'Standard YouTube License',
 879                 },
 880             }, {
 881                 'info_dict': {
 882                     'id': 'PUOgX5z9xZw',
 883                     'ext': 'mp4',
 884                     'title': 'teamPGP: Rocket League Noob Stream (grizzle)',
 885                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 886                     'duration': 7337,
 887                     'upload_date': '20150721',
 888                     'uploader': 'Beer Games Beer',
 889                     'uploader_id': 'beergamesbeer',
 890                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 891                     'license': 'Standard YouTube License',
 892                 },
 893             }, {
 894                 'info_dict': {
 895                     'id': 'teuwxikvS5k',
 896                     'ext': 'mp4',
 897                     'title': 'teamPGP: Rocket League Noob Stream (zim)',
 898                     'description': 'md5:dc7872fb300e143831327f1bae3af010',
 899                     'duration': 7334,
 900                     'upload_date': '20150721',
 901                     'uploader': 'Beer Games Beer',
 902                     'uploader_id': 'beergamesbeer',
 903                     'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/beergamesbeer',
 904                     'license': 'Standard YouTube License',
 905                 },
 906             }],
 907             'params': {
 908                 'skip_download': True,
 909             },
 910             'skip': 'This video is not available.',
 911         },
 912         {
 913             # Multifeed video with comma in title (see https://github.com/ytdl-org/youtube-dl/issues/8536)
 914             'url': 'https://www.youtube.com/watch?v=gVfLd0zydlo',
 915             'info_dict': {
 916                 'id': 'gVfLd0zydlo',
 917                 'title': 'DevConf.cz 2016 Day 2 Workshops 1 14:00 - 15:30',
 918             },
 919             'playlist_count': 2,
 920             'skip': 'Not multifeed anymore',
 921         },
 922         {
 923             'url': 'https://vid.plus/FlRa-iH7PGw',
 924             'only_matching': True,
 925         },
 926         {
 927             'url': 'https://zwearz.com/watch/9lWxNJF-ufM/electra-woman-dyna-girl-official-trailer-grace-helbig.html',
 928             'only_matching': True,
 929         },
 930         {
 931             # Title with JS-like syntax "};" (see https://github.com/ytdl-org/youtube-dl/issues/7468)
 932             # Also tests cut-off URL expansion in video description (see
 933             # https://github.com/ytdl-org/youtube-dl/issues/1892,
 934             # https://github.com/ytdl-org/youtube-dl/issues/8164)
 935             'url': 'https://www.youtube.com/watch?v=lsguqyKfVQg',
 936             'info_dict': {
 937                 'id': 'lsguqyKfVQg',
 938                 'ext': 'mp4',
 939                 'title': '{dark walk}; Loki/AC/Dishonored; collab w/Elflover21',
 940                 'alt_title': 'Dark Walk - Position Music',
 941                 'description': 'md5:8085699c11dc3f597ce0410b0dcbb34a',
 942                 'duration': 133,
 943                 'upload_date': '20151119',
 944                 'uploader_id': 'IronSoulElf',
 945                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/IronSoulElf',
 946                 'uploader': 'IronSoulElf',
 947                 'creator': 'Todd Haberman,  Daniel Law Heath and Aaron Kaplan',
 948                 'track': 'Dark Walk - Position Music',
 949                 'artist': 'Todd Haberman,  Daniel Law Heath and Aaron Kaplan',
 950                 'album': 'Position Music - Production Music Vol. 143 - Dark Walk',
 951             },
 952             'params': {
 953                 'skip_download': True,
 954             },
 955         },
 956         {
 957             # Tags with '};' (see https://github.com/ytdl-org/youtube-dl/issues/7468)
 958             'url': 'https://www.youtube.com/watch?v=Ms7iBXnlUO8',
 959             'only_matching': True,
 960         },
 961         {
 962             # Video with yt:stretch=17:0
 963             'url': 'https://www.youtube.com/watch?v=Q39EVAstoRM',
 964             'info_dict': {
 965                 'id': 'Q39EVAstoRM',
 966                 'ext': 'mp4',
 967                 'title': 'Clash Of Clans#14 Dicas De Ataque Para CV 4',
 968                 'description': 'md5:ee18a25c350637c8faff806845bddee9',
 969                 'upload_date': '20151107',
 970                 'uploader_id': 'UCCr7TALkRbo3EtFzETQF1LA',
 971                 'uploader': 'CH GAMER DROID',
 972             },
 973             'params': {
 974                 'skip_download': True,
 975             },
 976             'skip': 'This video does not exist.',
 977         },
 978         {
 979             # Video licensed under Creative Commons
 980             'url': 'https://www.youtube.com/watch?v=M4gD1WSo5mA',
 981             'info_dict': {
 982                 'id': 'M4gD1WSo5mA',
 983                 'ext': 'mp4',
 984                 'title': 'md5:e41008789470fc2533a3252216f1c1d1',
 985                 'description': 'md5:a677553cf0840649b731a3024aeff4cc',
 986                 'duration': 721,
 987                 'upload_date': '20150127',
 988                 'uploader_id': 'BerkmanCenter',
 989                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/BerkmanCenter',
 990                 'uploader': 'The Berkman Klein Center for Internet & Society',
 991                 'license': 'Creative Commons Attribution license (reuse allowed)',
 992             },
 993             'params': {
 994                 'skip_download': True,
 995             },
 996         },
 997         {
 998             # Channel-like uploader_url
 999             'url': 'https://www.youtube.com/watch?v=eQcmzGIKrzg',
1000             'info_dict': {
1001                 'id': 'eQcmzGIKrzg',
1002                 'ext': 'mp4',
1003                 'title': 'Democratic Socialism and Foreign Policy | Bernie Sanders',
1004                 'description': 'md5:dda0d780d5a6e120758d1711d062a867',
1005                 'duration': 4060,
1006                 'upload_date': '20151119',
1007                 'uploader': 'Bernie Sanders',
1008                 'uploader_id': 'UCH1dpzjCEiGAt8CXkryhkZg',
1009                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCH1dpzjCEiGAt8CXkryhkZg',
1010                 'license': 'Creative Commons Attribution license (reuse allowed)',
1011             },
1012             'params': {
1013                 'skip_download': True,
1014             },
1015         },
1016         {
1017             'url': 'https://www.youtube.com/watch?feature=player_embedded&amp;amp;v=V36LpHqtcDY',
1018             'only_matching': True,
1019         },
1020         {
1021             # YouTube Red paid video (https://github.com/ytdl-org/youtube-dl/issues/10059)
1022             'url': 'https://www.youtube.com/watch?v=i1Ko8UG-Tdo',
1023             'only_matching': True,
1024         },
1025         {
1026             # Rental video preview
1027             'url': 'https://www.youtube.com/watch?v=yYr8q0y5Jfg',
1028             'info_dict': {
1029                 'id': 'uGpuVWrhIzE',
1030                 'ext': 'mp4',
1031                 'title': 'Piku - Trailer',
1032                 'description': 'md5:c36bd60c3fd6f1954086c083c72092eb',
1033                 'upload_date': '20150811',
1034                 'uploader': 'FlixMatrix',
1035                 'uploader_id': 'FlixMatrixKaravan',
1036                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/FlixMatrixKaravan',
1037                 'license': 'Standard YouTube License',
1038             },
1039             'params': {
1040                 'skip_download': True,
1041             },
1042             'skip': 'This video is not available.',
1043         },
1044         {
1045             # YouTube Red video with episode data
1046             'url': 'https://www.youtube.com/watch?v=iqKdEhx-dD4',
1047             'info_dict': {
1048                 'id': 'iqKdEhx-dD4',
1049                 'ext': 'mp4',
1050                 'title': 'Isolation - Mind Field (Ep 1)',
1051                 'description': 'md5:46a29be4ceffa65b92d277b93f463c0f',
1052                 'duration': 2085,
1053                 'upload_date': '20170118',
1054                 'uploader': 'Vsauce',
1055                 'uploader_id': 'Vsauce',
1056                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/Vsauce',
1057                 'series': 'Mind Field',
1058                 'season_number': 1,
1059                 'episode_number': 1,
1060             },
1061             'params': {
1062                 'skip_download': True,
1063             },
1064             'expected_warnings': [
1065                 'Skipping DASH manifest',
1066             ],
1067         },
1068         {
1069             # The following content has been identified by the YouTube community
1070             # as inappropriate or offensive to some audiences.
1071             'url': 'https://www.youtube.com/watch?v=6SJNVb0GnPI',
1072             'info_dict': {
1073                 'id': '6SJNVb0GnPI',
1074                 'ext': 'mp4',
1075                 'title': 'Race Differences in Intelligence',
1076                 'description': 'md5:5d161533167390427a1f8ee89a1fc6f1',
1077                 'duration': 965,
1078                 'upload_date': '20140124',
1079                 'uploader': 'New Century Foundation',
1080                 'uploader_id': 'UCEJYpZGqgUob0zVVEaLhvVg',
1081                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCEJYpZGqgUob0zVVEaLhvVg',
1082             },
1083             'params': {
1084                 'skip_download': True,
1085             },
1086         },
1087         {
1088             # itag 212
1089             'url': '1t24XAntNCY',
1090             'only_matching': True,
1091         },
1092         {
1093             # geo restricted to JP
1094             'url': 'sJL6WA-aGkQ',
1095             'only_matching': True,
1096         },
1097         {
1098             'url': 'https://www.youtube.com/watch?v=MuAGGZNfUkU&list=RDMM',
1099             'only_matching': True,
1100         },
1101         {
1102             'url': 'https://invidio.us/watch?v=BaW_jenozKc',
1103             'only_matching': True,
1104         },
1105         {
1106             # DRM protected
1107             'url': 'https://www.youtube.com/watch?v=s7_qI6_mIXc',
1108             'only_matching': True,
1109         },
1110         {
1111             # Video with unsupported adaptive stream type formats
1112             'url': 'https://www.youtube.com/watch?v=Z4Vy8R84T1U',
1113             'info_dict': {
1114                 'id': 'Z4Vy8R84T1U',
1115                 'ext': 'mp4',
1116                 'title': 'saman SMAN 53 Jakarta(Sancety) opening COFFEE4th at SMAN 53 Jakarta',
1117                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
1118                 'duration': 433,
1119                 'upload_date': '20130923',
1120                 'uploader': 'Amelia Putri Harwita',
1121                 'uploader_id': 'UCpOxM49HJxmC1qCalXyB3_Q',
1122                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCpOxM49HJxmC1qCalXyB3_Q',
1123                 'formats': 'maxcount:10',
1124             },
1125             'params': {
1126                 'skip_download': True,
1127                 'youtube_include_dash_manifest': False,
1128             },
1129         },
1130         {
1131             # Youtube Music Auto-generated description
1132             'url': 'https://music.youtube.com/watch?v=MgNrAu2pzNs',
1133             'info_dict': {
1134                 'id': 'MgNrAu2pzNs',
1135                 'ext': 'mp4',
1136                 'title': 'Voyeur Girl',
1137                 'description': 'md5:7ae382a65843d6df2685993e90a8628f',
1138                 'upload_date': '20190312',
1139                 'uploader': 'Various Artists - Topic',
1140                 'uploader_id': 'UCVWKBi1ELZn0QX2CBLSkiyw',
1141                 'artist': 'Stephen',
1142                 'track': 'Voyeur Girl',
1143                 'album': 'it\'s too much love to know my dear',
1144                 'release_date': '20190313',
1145                 'release_year': 2019,
1146             },
1147             'params': {
1148                 'skip_download': True,
1149             },
1150         },
1151         {
1152             # Youtube Music Auto-generated description
1153             # Retrieve 'artist' field from 'Artist:' in video description
1154             # when it is present on youtube music video
1155             'url': 'https://www.youtube.com/watch?v=k0jLE7tTwjY',
1156             'info_dict': {
1157                 'id': 'k0jLE7tTwjY',
1158                 'ext': 'mp4',
1159                 'title': 'Latch Feat. Sam Smith',
1160                 'description': 'md5:3cb1e8101a7c85fcba9b4fb41b951335',
1161                 'upload_date': '20150110',
1162                 'uploader': 'Various Artists - Topic',
1163                 'uploader_id': 'UCNkEcmYdjrH4RqtNgh7BZ9w',
1164                 'artist': 'Disclosure',
1165                 'track': 'Latch Feat. Sam Smith',
1166                 'album': 'Latch Featuring Sam Smith',
1167                 'release_date': '20121008',
1168                 'release_year': 2012,
1169             },
1170             'params': {
1171                 'skip_download': True,
1172             },
1173         },
1174         {
1175             # Youtube Music Auto-generated description
1176             # handle multiple artists on youtube music video
1177             'url': 'https://www.youtube.com/watch?v=74qn0eJSjpA',
1178             'info_dict': {
1179                 'id': '74qn0eJSjpA',
1180                 'ext': 'mp4',
1181                 'title': 'Eastside',
1182                 'description': 'md5:290516bb73dcbfab0dcc4efe6c3de5f2',
1183                 'upload_date': '20180710',
1184                 'uploader': 'Benny Blanco - Topic',
1185                 'uploader_id': 'UCzqz_ksRu_WkIzmivMdIS7A',
1186                 'artist': 'benny blanco, Halsey, Khalid',
1187                 'track': 'Eastside',
1188                 'album': 'Eastside',
1189                 'release_date': '20180713',
1190                 'release_year': 2018,
1191             },
1192             'params': {
1193                 'skip_download': True,
1194             },
1195         },
1196         {
1197             # Youtube Music Auto-generated description
1198             # handle youtube music video with release_year and no release_date
1199             'url': 'https://www.youtube.com/watch?v=-hcAI0g-f5M',
1200             'info_dict': {
1201                 'id': '-hcAI0g-f5M',
1202                 'ext': 'mp4',
1203                 'title': 'Put It On Me',
1204                 'description': 'md5:93c55acc682ae7b0c668f2e34e1c069e',
1205                 'upload_date': '20180426',
1206                 'uploader': 'Matt Maeson - Topic',
1207                 'uploader_id': 'UCnEkIGqtGcQMLk73Kp-Q5LQ',
1208                 'artist': 'Matt Maeson',
1209                 'track': 'Put It On Me',
1210                 'album': 'The Hearse',
1211                 'release_date': None,
1212                 'release_year': 2018,
1213             },
1214             'params': {
1215                 'skip_download': True,
1216             },
1217         },
1218     ]
1219
1220     def __init__(self, *args, **kwargs):
1221         super(YoutubeIE, self).__init__(*args, **kwargs)
1222         self._player_cache = {}
1223
1224     def report_video_info_webpage_download(self, video_id):
1225         """Report attempt to download video info webpage."""
1226         self.to_screen('%s: Downloading video info webpage' % video_id)
1227
1228     def report_information_extraction(self, video_id):
1229         """Report attempt to extract video information."""
1230         self.to_screen('%s: Extracting video information' % video_id)
1231
1232     def report_unavailable_format(self, video_id, format):
1233         """Report extracted video URL."""
1234         self.to_screen('%s: Format %s not available' % (video_id, format))
1235
1236     def report_rtmp_download(self):
1237         """Indicate the download will use the RTMP protocol."""
1238         self.to_screen('RTMP download detected')
1239
1240     def _signature_cache_id(self, example_sig):
1241         """ Return a string representation of a signature """
1242         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
1243
1244     def _extract_signature_function(self, video_id, player_url, example_sig):
1245         id_m = re.match(
1246             r'.*?-(?P<id>[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player(?:-new)?|(?:/[a-z]{2,3}_[A-Z]{2})?/base)?\.(?P<ext>[a-z]+)$',
1247             player_url)
1248         if not id_m:
1249             raise ExtractorError('Cannot identify player %r' % player_url)
1250         player_type = id_m.group('ext')
1251         player_id = id_m.group('id')
1252
1253         # Read from filesystem cache
1254         func_id = '%s_%s_%s' % (
1255             player_type, player_id, self._signature_cache_id(example_sig))
1256         assert os.path.basename(func_id) == func_id
1257
1258         cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
1259         if cache_spec is not None:
1260             return lambda s: ''.join(s[i] for i in cache_spec)
1261
1262         download_note = (
1263             'Downloading player %s' % player_url
1264             if self._downloader.params.get('verbose') else
1265             'Downloading %s player %s' % (player_type, player_id)
1266         )
1267         if player_type == 'js':
1268             code = self._download_webpage(
1269                 player_url, video_id,
1270                 note=download_note,
1271                 errnote='Download of %s failed' % player_url)
1272             res = self._parse_sig_js(code)
1273         elif player_type == 'swf':
1274             urlh = self._request_webpage(
1275                 player_url, video_id,
1276                 note=download_note,
1277                 errnote='Download of %s failed' % player_url)
1278             code = urlh.read()
1279             res = self._parse_sig_swf(code)
1280         else:
1281             assert False, 'Invalid player type %r' % player_type
1282
1283         test_string = ''.join(map(compat_chr, range(len(example_sig))))
1284         cache_res = res(test_string)
1285         cache_spec = [ord(c) for c in cache_res]
1286
1287         self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
1288         return res
1289
1290     def _print_sig_code(self, func, example_sig):
1291         def gen_sig_code(idxs):
1292             def _genslice(start, end, step):
1293                 starts = '' if start == 0 else str(start)
1294                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
1295                 steps = '' if step == 1 else (':%d' % step)
1296                 return 's[%s%s%s]' % (starts, ends, steps)
1297
1298             step = None
1299             # Quelch pyflakes warnings - start will be set when step is set
1300             start = '(Never used)'
1301             for i, prev in zip(idxs[1:], idxs[:-1]):
1302                 if step is not None:
1303                     if i - prev == step:
1304                         continue
1305                     yield _genslice(start, prev, step)
1306                     step = None
1307                     continue
1308                 if i - prev in [-1, 1]:
1309                     step = i - prev
1310                     start = prev
1311                     continue
1312                 else:
1313                     yield 's[%d]' % prev
1314             if step is None:
1315                 yield 's[%d]' % i
1316             else:
1317                 yield _genslice(start, i, step)
1318
1319         test_string = ''.join(map(compat_chr, range(len(example_sig))))
1320         cache_res = func(test_string)
1321         cache_spec = [ord(c) for c in cache_res]
1322         expr_code = ' + '.join(gen_sig_code(cache_spec))
1323         signature_id_tuple = '(%s)' % (
1324             ', '.join(compat_str(len(p)) for p in example_sig.split('.')))
1325         code = ('if tuple(len(p) for p in s.split(\'.\')) == %s:\n'
1326                 '    return %s\n') % (signature_id_tuple, expr_code)
1327         self.to_screen('Extracted signature function:\n' + code)
1328
1329     def _parse_sig_js(self, jscode):
1330         funcname = self._search_regex(
1331             (r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1332              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1333              r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)',
1334              # Obsolete patterns
1335              r'(["\'])signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1336              r'\.sig\|\|(?P<sig>[a-zA-Z0-9$]+)\(',
1337              r'yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1338              r'\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1339              r'\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1340              r'\bc\s*&&\s*a\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1341              r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(',
1342              r'\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\('),
1343             jscode, 'Initial JS player signature function name', group='sig')
1344
1345         jsi = JSInterpreter(jscode)
1346         initial_function = jsi.extract_function(funcname)
1347         return lambda s: initial_function([s])
1348
1349     def _parse_sig_swf(self, file_contents):
1350         swfi = SWFInterpreter(file_contents)
1351         TARGET_CLASSNAME = 'SignatureDecipher'
1352         searched_class = swfi.extract_class(TARGET_CLASSNAME)
1353         initial_function = swfi.extract_function(searched_class, 'decipher')
1354         return lambda s: initial_function([s])
1355
1356     def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
1357         """Turn the encrypted s field into a working signature"""
1358
1359         if player_url is None:
1360             raise ExtractorError('Cannot decrypt signature without player_url')
1361
1362         if player_url.startswith('//'):
1363             player_url = 'https:' + player_url
1364         elif not re.match(r'https?://', player_url):
1365             player_url = compat_urlparse.urljoin(
1366                 'https://www.youtube.com', player_url)
1367         try:
1368             player_id = (player_url, self._signature_cache_id(s))
1369             if player_id not in self._player_cache:
1370                 func = self._extract_signature_function(
1371                     video_id, player_url, s
1372                 )
1373                 self._player_cache[player_id] = func
1374             func = self._player_cache[player_id]
1375             if self._downloader.params.get('youtube_print_sig_code'):
1376                 self._print_sig_code(func, s)
1377             return func(s)
1378         except Exception as e:
1379             tb = traceback.format_exc()
1380             raise ExtractorError(
1381                 'Signature extraction failed: ' + tb, cause=e)
1382
1383     def _get_subtitles(self, video_id, webpage):
1384         try:
1385             subs_doc = self._download_xml(
1386                 'https://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
1387                 video_id, note=False)
1388         except ExtractorError as err:
1389             self._downloader.report_warning('unable to download video subtitles: %s' % error_to_compat_str(err))
1390             return {}
1391
1392         sub_lang_list = {}
1393         for track in subs_doc.findall('track'):
1394             lang = track.attrib['lang_code']
1395             if lang in sub_lang_list:
1396                 continue
1397             sub_formats = []
1398             for ext in self._SUBTITLE_FORMATS:
1399                 params = compat_urllib_parse_urlencode({
1400                     'lang': lang,
1401                     'v': video_id,
1402                     'fmt': ext,
1403                     'name': track.attrib['name'].encode('utf-8'),
1404                 })
1405                 sub_formats.append({
1406                     'url': 'https://www.youtube.com/api/timedtext?' + params,
1407                     'ext': ext,
1408                 })
1409             sub_lang_list[lang] = sub_formats
1410         if not sub_lang_list:
1411             self._downloader.report_warning('video doesn\'t have subtitles')
1412             return {}
1413         return sub_lang_list
1414
1415     def _get_ytplayer_config(self, video_id, webpage):
1416         patterns = (
1417             # User data may contain arbitrary character sequences that may affect
1418             # JSON extraction with regex, e.g. when '};' is contained the second
1419             # regex won't capture the whole JSON. Yet working around by trying more
1420             # concrete regex first keeping in mind proper quoted string handling
1421             # to be implemented in future that will replace this workaround (see
1422             # https://github.com/ytdl-org/youtube-dl/issues/7468,
1423             # https://github.com/ytdl-org/youtube-dl/pull/7599)
1424             r';ytplayer\.config\s*=\s*({.+?});ytplayer',
1425             r';ytplayer\.config\s*=\s*({.+?});',
1426         )
1427         config = self._search_regex(
1428             patterns, webpage, 'ytplayer.config', default=None)
1429         if config:
1430             return self._parse_json(
1431                 uppercase_escape(config), video_id, fatal=False)
1432
1433     def _get_automatic_captions(self, video_id, webpage):
1434         """We need the webpage for getting the captions url, pass it as an
1435            argument to speed up the process."""
1436         self.to_screen('%s: Looking for automatic captions' % video_id)
1437         player_config = self._get_ytplayer_config(video_id, webpage)
1438         err_msg = 'Couldn\'t find automatic captions for %s' % video_id
1439         if not player_config:
1440             self._downloader.report_warning(err_msg)
1441             return {}
1442         try:
1443             args = player_config['args']
1444             caption_url = args.get('ttsurl')
1445             if caption_url:
1446                 timestamp = args['timestamp']
1447                 # We get the available subtitles
1448                 list_params = compat_urllib_parse_urlencode({
1449                     'type': 'list',
1450                     'tlangs': 1,
1451                     'asrs': 1,
1452                 })
1453                 list_url = caption_url + '&' + list_params
1454                 caption_list = self._download_xml(list_url, video_id)
1455                 original_lang_node = caption_list.find('track')
1456                 if original_lang_node is None:
1457                     self._downloader.report_warning('Video doesn\'t have automatic captions')
1458                     return {}
1459                 original_lang = original_lang_node.attrib['lang_code']
1460                 caption_kind = original_lang_node.attrib.get('kind', '')
1461
1462                 sub_lang_list = {}
1463                 for lang_node in caption_list.findall('target'):
1464                     sub_lang = lang_node.attrib['lang_code']
1465                     sub_formats = []
1466                     for ext in self._SUBTITLE_FORMATS:
1467                         params = compat_urllib_parse_urlencode({
1468                             'lang': original_lang,
1469                             'tlang': sub_lang,
1470                             'fmt': ext,
1471                             'ts': timestamp,
1472                             'kind': caption_kind,
1473                         })
1474                         sub_formats.append({
1475                             'url': caption_url + '&' + params,
1476                             'ext': ext,
1477                         })
1478                     sub_lang_list[sub_lang] = sub_formats
1479                 return sub_lang_list
1480
1481             def make_captions(sub_url, sub_langs):
1482                 parsed_sub_url = compat_urllib_parse_urlparse(sub_url)
1483                 caption_qs = compat_parse_qs(parsed_sub_url.query)
1484                 captions = {}
1485                 for sub_lang in sub_langs:
1486                     sub_formats = []
1487                     for ext in self._SUBTITLE_FORMATS:
1488                         caption_qs.update({
1489                             'tlang': [sub_lang],
1490                             'fmt': [ext],
1491                         })
1492                         sub_url = compat_urlparse.urlunparse(parsed_sub_url._replace(
1493                             query=compat_urllib_parse_urlencode(caption_qs, True)))
1494                         sub_formats.append({
1495                             'url': sub_url,
1496                             'ext': ext,
1497                         })
1498                     captions[sub_lang] = sub_formats
1499                 return captions
1500
1501             # New captions format as of 22.06.2017
1502             player_response = args.get('player_response')
1503             if player_response and isinstance(player_response, compat_str):
1504                 player_response = self._parse_json(
1505                     player_response, video_id, fatal=False)
1506                 if player_response:
1507                     renderer = player_response['captions']['playerCaptionsTracklistRenderer']
1508                     base_url = renderer['captionTracks'][0]['baseUrl']
1509                     sub_lang_list = []
1510                     for lang in renderer['translationLanguages']:
1511                         lang_code = lang.get('languageCode')
1512                         if lang_code:
1513                             sub_lang_list.append(lang_code)
1514                     return make_captions(base_url, sub_lang_list)
1515
1516             # Some videos don't provide ttsurl but rather caption_tracks and
1517             # caption_translation_languages (e.g. 20LmZk1hakA)
1518             # Does not used anymore as of 22.06.2017
1519             caption_tracks = args['caption_tracks']
1520             caption_translation_languages = args['caption_translation_languages']
1521             caption_url = compat_parse_qs(caption_tracks.split(',')[0])['u'][0]
1522             sub_lang_list = []
1523             for lang in caption_translation_languages.split(','):
1524                 lang_qs = compat_parse_qs(compat_urllib_parse_unquote_plus(lang))
1525                 sub_lang = lang_qs.get('lc', [None])[0]
1526                 if sub_lang:
1527                     sub_lang_list.append(sub_lang)
1528             return make_captions(caption_url, sub_lang_list)
1529         # An extractor error can be raise by the download process if there are
1530         # no automatic captions but there are subtitles
1531         except (KeyError, IndexError, ExtractorError):
1532             self._downloader.report_warning(err_msg)
1533             return {}
1534
1535     def _mark_watched(self, video_id, video_info, player_response):
1536         playback_url = url_or_none(try_get(
1537             player_response,
1538             lambda x: x['playbackTracking']['videostatsPlaybackUrl']['baseUrl']) or try_get(
1539             video_info, lambda x: x['videostats_playback_base_url'][0]))
1540         if not playback_url:
1541             return
1542         parsed_playback_url = compat_urlparse.urlparse(playback_url)
1543         qs = compat_urlparse.parse_qs(parsed_playback_url.query)
1544
1545         # cpn generation algorithm is reverse engineered from base.js.
1546         # In fact it works even with dummy cpn.
1547         CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
1548         cpn = ''.join((CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16)))
1549
1550         qs.update({
1551             'ver': ['2'],
1552             'cpn': [cpn],
1553         })
1554         playback_url = compat_urlparse.urlunparse(
1555             parsed_playback_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
1556
1557         self._download_webpage(
1558             playback_url, video_id, 'Marking watched',
1559             'Unable to mark watched', fatal=False)
1560
1561     @staticmethod
1562     def _extract_urls(webpage):
1563         # Embedded YouTube player
1564         entries = [
1565             unescapeHTML(mobj.group('url'))
1566             for mobj in re.finditer(r'''(?x)
1567             (?:
1568                 <iframe[^>]+?src=|
1569                 data-video-url=|
1570                 <embed[^>]+?src=|
1571                 embedSWF\(?:\s*|
1572                 <object[^>]+data=|
1573                 new\s+SWFObject\(
1574             )
1575             (["\'])
1576                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
1577                 (?:embed|v|p)/[0-9A-Za-z_-]{11}.*?)
1578             \1''', webpage)]
1579
1580         # lazyYT YouTube embed
1581         entries.extend(list(map(
1582             unescapeHTML,
1583             re.findall(r'class="lazyYT" data-youtube-id="([^"]+)"', webpage))))
1584
1585         # Wordpress "YouTube Video Importer" plugin
1586         matches = re.findall(r'''(?x)<div[^>]+
1587             class=(?P<q1>[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+
1588             data-video_id=(?P<q2>[\'"])([^\'"]+)(?P=q2)''', webpage)
1589         entries.extend(m[-1] for m in matches)
1590
1591         return entries
1592
1593     @staticmethod
1594     def _extract_url(webpage):
1595         urls = YoutubeIE._extract_urls(webpage)
1596         return urls[0] if urls else None
1597
1598     @classmethod
1599     def extract_id(cls, url):
1600         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
1601         if mobj is None:
1602             raise ExtractorError('Invalid URL: %s' % url)
1603         video_id = mobj.group(2)
1604         return video_id
1605
1606     @staticmethod
1607     def _extract_chapters(description, duration):
1608         if not description:
1609             return None
1610         chapter_lines = re.findall(
1611             r'(?:^|<br\s*/>)([^<]*<a[^>]+onclick=["\']yt\.www\.watch\.player\.seekTo[^>]+>(\d{1,2}:\d{1,2}(?::\d{1,2})?)</a>[^>]*)(?=$|<br\s*/>)',
1612             description)
1613         if not chapter_lines:
1614             return None
1615         chapters = []
1616         for next_num, (chapter_line, time_point) in enumerate(
1617                 chapter_lines, start=1):
1618             start_time = parse_duration(time_point)
1619             if start_time is None:
1620                 continue
1621             if start_time > duration:
1622                 break
1623             end_time = (duration if next_num == len(chapter_lines)
1624                         else parse_duration(chapter_lines[next_num][1]))
1625             if end_time is None:
1626                 continue
1627             if end_time > duration:
1628                 end_time = duration
1629             if start_time > end_time:
1630                 break
1631             chapter_title = re.sub(
1632                 r'<a[^>]+>[^<]+</a>', '', chapter_line).strip(' \t-')
1633             chapter_title = re.sub(r'\s+', ' ', chapter_title)
1634             chapters.append({
1635                 'start_time': start_time,
1636                 'end_time': end_time,
1637                 'title': chapter_title,
1638             })
1639         return chapters
1640
1641     def _real_extract(self, url):
1642         url, smuggled_data = unsmuggle_url(url, {})
1643
1644         proto = (
1645             'http' if self._downloader.params.get('prefer_insecure', False)
1646             else 'https')
1647
1648         start_time = None
1649         end_time = None
1650         parsed_url = compat_urllib_parse_urlparse(url)
1651         for component in [parsed_url.fragment, parsed_url.query]:
1652             query = compat_parse_qs(component)
1653             if start_time is None and 't' in query:
1654                 start_time = parse_duration(query['t'][0])
1655             if start_time is None and 'start' in query:
1656                 start_time = parse_duration(query['start'][0])
1657             if end_time is None and 'end' in query:
1658                 end_time = parse_duration(query['end'][0])
1659
1660         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
1661         mobj = re.search(self._NEXT_URL_RE, url)
1662         if mobj:
1663             url = proto + '://www.youtube.com/' + compat_urllib_parse_unquote(mobj.group(1)).lstrip('/')
1664         video_id = self.extract_id(url)
1665
1666         # Get video webpage
1667         url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1&bpctr=9999999999' % video_id
1668         video_webpage = self._download_webpage(url, video_id)
1669
1670         # Attempt to extract SWF player URL
1671         mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
1672         if mobj is not None:
1673             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
1674         else:
1675             player_url = None
1676
1677         dash_mpds = []
1678
1679         def add_dash_mpd(video_info):
1680             dash_mpd = video_info.get('dashmpd')
1681             if dash_mpd and dash_mpd[0] not in dash_mpds:
1682                 dash_mpds.append(dash_mpd[0])
1683
1684         def add_dash_mpd_pr(pl_response):
1685             dash_mpd = url_or_none(try_get(
1686                 pl_response, lambda x: x['streamingData']['dashManifestUrl'],
1687                 compat_str))
1688             if dash_mpd and dash_mpd not in dash_mpds:
1689                 dash_mpds.append(dash_mpd)
1690
1691         is_live = None
1692         view_count = None
1693
1694         def extract_view_count(v_info):
1695             return int_or_none(try_get(v_info, lambda x: x['view_count'][0]))
1696
1697         def extract_token(v_info):
1698             return dict_get(v_info, ('account_playback_token', 'accountPlaybackToken', 'token'))
1699
1700         def extract_player_response(player_response, video_id):
1701             pl_response = str_or_none(player_response)
1702             if not pl_response:
1703                 return
1704             pl_response = self._parse_json(pl_response, video_id, fatal=False)
1705             if isinstance(pl_response, dict):
1706                 add_dash_mpd_pr(pl_response)
1707                 return pl_response
1708
1709         player_response = {}
1710
1711         # Get video info
1712         embed_webpage = None
1713         if re.search(r'player-age-gate-content">', video_webpage) is not None:
1714             age_gate = True
1715             # We simulate the access to the video from www.youtube.com/v/{video_id}
1716             # this can be viewed without login into Youtube
1717             url = proto + '://www.youtube.com/embed/%s' % video_id
1718             embed_webpage = self._download_webpage(url, video_id, 'Downloading embed webpage')
1719             data = compat_urllib_parse_urlencode({
1720                 'video_id': video_id,
1721                 'eurl': 'https://youtube.googleapis.com/v/' + video_id,
1722                 'sts': self._search_regex(
1723                     r'"sts"\s*:\s*(\d+)', embed_webpage, 'sts', default=''),
1724             })
1725             video_info_url = proto + '://www.youtube.com/get_video_info?' + data
1726             video_info_webpage = self._download_webpage(
1727                 video_info_url, video_id,
1728                 note='Refetching age-gated info webpage',
1729                 errnote='unable to download video info webpage')
1730             video_info = compat_parse_qs(video_info_webpage)
1731             pl_response = video_info.get('player_response', [None])[0]
1732             player_response = extract_player_response(pl_response, video_id)
1733             add_dash_mpd(video_info)
1734             view_count = extract_view_count(video_info)
1735         else:
1736             age_gate = False
1737             video_info = None
1738             sts = None
1739             # Try looking directly into the video webpage
1740             ytplayer_config = self._get_ytplayer_config(video_id, video_webpage)
1741             if ytplayer_config:
1742                 args = ytplayer_config['args']
1743                 if args.get('url_encoded_fmt_stream_map') or args.get('hlsvp'):
1744                     # Convert to the same format returned by compat_parse_qs
1745                     video_info = dict((k, [v]) for k, v in args.items())
1746                     add_dash_mpd(video_info)
1747                 # Rental video is not rented but preview is available (e.g.
1748                 # https://www.youtube.com/watch?v=yYr8q0y5Jfg,
1749                 # https://github.com/ytdl-org/youtube-dl/issues/10532)
1750                 if not video_info and args.get('ypc_vid'):
1751                     return self.url_result(
1752                         args['ypc_vid'], YoutubeIE.ie_key(), video_id=args['ypc_vid'])
1753                 if args.get('livestream') == '1' or args.get('live_playback') == 1:
1754                     is_live = True
1755                 sts = ytplayer_config.get('sts')
1756                 if not player_response:
1757                     player_response = extract_player_response(args.get('player_response'), video_id)
1758             if not video_info or self._downloader.params.get('youtube_include_dash_manifest', True):
1759                 add_dash_mpd_pr(player_response)
1760                 # We also try looking in get_video_info since it may contain different dashmpd
1761                 # URL that points to a DASH manifest with possibly different itag set (some itags
1762                 # are missing from DASH manifest pointed by webpage's dashmpd, some - from DASH
1763                 # manifest pointed by get_video_info's dashmpd).
1764                 # The general idea is to take a union of itags of both DASH manifests (for example
1765                 # video with such 'manifest behavior' see https://github.com/ytdl-org/youtube-dl/issues/6093)
1766                 self.report_video_info_webpage_download(video_id)
1767                 for el in ('embedded', 'detailpage', 'vevo', ''):
1768                     query = {
1769                         'video_id': video_id,
1770                         'ps': 'default',
1771                         'eurl': '',
1772                         'gl': 'US',
1773                         'hl': 'en',
1774                     }
1775                     if el:
1776                         query['el'] = el
1777                     if sts:
1778                         query['sts'] = sts
1779                     video_info_webpage = self._download_webpage(
1780                         '%s://www.youtube.com/get_video_info' % proto,
1781                         video_id, note=False,
1782                         errnote='unable to download video info webpage',
1783                         fatal=False, query=query)
1784                     if not video_info_webpage:
1785                         continue
1786                     get_video_info = compat_parse_qs(video_info_webpage)
1787                     if not player_response:
1788                         pl_response = get_video_info.get('player_response', [None])[0]
1789                         player_response = extract_player_response(pl_response, video_id)
1790                     add_dash_mpd(get_video_info)
1791                     if view_count is None:
1792                         view_count = extract_view_count(get_video_info)
1793                     if not video_info:
1794                         video_info = get_video_info
1795                     get_token = extract_token(get_video_info)
1796                     if get_token:
1797                         # Different get_video_info requests may report different results, e.g.
1798                         # some may report video unavailability, but some may serve it without
1799                         # any complaint (see https://github.com/ytdl-org/youtube-dl/issues/7362,
1800                         # the original webpage as well as el=info and el=embedded get_video_info
1801                         # requests report video unavailability due to geo restriction while
1802                         # el=detailpage succeeds and returns valid data). This is probably
1803                         # due to YouTube measures against IP ranges of hosting providers.
1804                         # Working around by preferring the first succeeded video_info containing
1805                         # the token if no such video_info yet was found.
1806                         token = extract_token(video_info)
1807                         if not token:
1808                             video_info = get_video_info
1809                         break
1810
1811         def extract_unavailable_message():
1812             messages = []
1813             for tag, kind in (('h1', 'message'), ('div', 'submessage')):
1814                 msg = self._html_search_regex(
1815                     r'(?s)<{tag}[^>]+id=["\']unavailable-{kind}["\'][^>]*>(.+?)</{tag}>'.format(tag=tag, kind=kind),
1816                     video_webpage, 'unavailable %s' % kind, default=None)
1817                 if msg:
1818                     messages.append(msg)
1819             if messages:
1820                 return '\n'.join(messages)
1821
1822         if not video_info:
1823             unavailable_message = extract_unavailable_message()
1824             if not unavailable_message:
1825                 unavailable_message = 'Unable to extract video data'
1826             raise ExtractorError(
1827                 'YouTube said: %s' % unavailable_message, expected=True, video_id=video_id)
1828
1829         video_details = try_get(
1830             player_response, lambda x: x['videoDetails'], dict) or {}
1831
1832         video_title = video_info.get('title', [None])[0] or video_details.get('title')
1833         if not video_title:
1834             self._downloader.report_warning('Unable to extract video title')
1835             video_title = '_'
1836
1837         description_original = video_description = get_element_by_id("eow-description", video_webpage)
1838         if video_description:
1839
1840             def replace_url(m):
1841                 redir_url = compat_urlparse.urljoin(url, m.group(1))
1842                 parsed_redir_url = compat_urllib_parse_urlparse(redir_url)
1843                 if re.search(r'^(?:www\.)?(?:youtube(?:-nocookie)?\.com|youtu\.be)$', parsed_redir_url.netloc) and parsed_redir_url.path == '/redirect':
1844                     qs = compat_parse_qs(parsed_redir_url.query)
1845                     q = qs.get('q')
1846                     if q and q[0]:
1847                         return q[0]
1848                 return redir_url
1849
1850             description_original = video_description = re.sub(r'''(?x)
1851                 <a\s+
1852                     (?:[a-zA-Z-]+="[^"]*"\s+)*?
1853                     (?:title|href)="([^"]+)"\s+
1854                     (?:[a-zA-Z-]+="[^"]*"\s+)*?
1855                     class="[^"]*"[^>]*>
1856                 [^<]+\.{3}\s*
1857                 </a>
1858             ''', replace_url, video_description)
1859             video_description = clean_html(video_description)
1860         else:
1861             video_description = self._html_search_meta('description', video_webpage) or video_details.get('shortDescription')
1862
1863         if not smuggled_data.get('force_singlefeed', False):
1864             if not self._downloader.params.get('noplaylist'):
1865                 multifeed_metadata_list = try_get(
1866                     player_response,
1867                     lambda x: x['multicamera']['playerLegacyMulticameraRenderer']['metadataList'],
1868                     compat_str) or try_get(
1869                     video_info, lambda x: x['multifeed_metadata_list'][0], compat_str)
1870                 if multifeed_metadata_list:
1871                     entries = []
1872                     feed_ids = []
1873                     for feed in multifeed_metadata_list.split(','):
1874                         # Unquote should take place before split on comma (,) since textual
1875                         # fields may contain comma as well (see
1876                         # https://github.com/ytdl-org/youtube-dl/issues/8536)
1877                         feed_data = compat_parse_qs(compat_urllib_parse_unquote_plus(feed))
1878                         entries.append({
1879                             '_type': 'url_transparent',
1880                             'ie_key': 'Youtube',
1881                             'url': smuggle_url(
1882                                 '%s://www.youtube.com/watch?v=%s' % (proto, feed_data['id'][0]),
1883                                 {'force_singlefeed': True}),
1884                             'title': '%s (%s)' % (video_title, feed_data['title'][0]),
1885                         })
1886                         feed_ids.append(feed_data['id'][0])
1887                     self.to_screen(
1888                         'Downloading multifeed video (%s) - add --no-playlist to just download video %s'
1889                         % (', '.join(feed_ids), video_id))
1890                     return self.playlist_result(entries, video_id, video_title, video_description)
1891             else:
1892                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
1893
1894         if view_count is None:
1895             view_count = extract_view_count(video_info)
1896         if view_count is None and video_details:
1897             view_count = int_or_none(video_details.get('viewCount'))
1898
1899         if is_live is None:
1900             is_live = bool_or_none(video_details.get('isLive'))
1901
1902         # Check for "rental" videos
1903         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
1904             raise ExtractorError('"rental" videos not supported. See https://github.com/ytdl-org/youtube-dl/issues/359 for more information.', expected=True)
1905
1906         def _extract_filesize(media_url):
1907             return int_or_none(self._search_regex(
1908                 r'\bclen[=/](\d+)', media_url, 'filesize', default=None))
1909
1910         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
1911             self.report_rtmp_download()
1912             formats = [{
1913                 'format_id': '_rtmp',
1914                 'protocol': 'rtmp',
1915                 'url': video_info['conn'][0],
1916                 'player_url': player_url,
1917             }]
1918         elif not is_live and (len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1):
1919             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
1920             if 'rtmpe%3Dyes' in encoded_url_map:
1921                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/ytdl-org/youtube-dl/issues/343 for more information.', expected=True)
1922             formats_spec = {}
1923             fmt_list = video_info.get('fmt_list', [''])[0]
1924             if fmt_list:
1925                 for fmt in fmt_list.split(','):
1926                     spec = fmt.split('/')
1927                     if len(spec) > 1:
1928                         width_height = spec[1].split('x')
1929                         if len(width_height) == 2:
1930                             formats_spec[spec[0]] = {
1931                                 'resolution': spec[1],
1932                                 'width': int_or_none(width_height[0]),
1933                                 'height': int_or_none(width_height[1]),
1934                             }
1935             q = qualities(['small', 'medium', 'hd720'])
1936             streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list)
1937             if streaming_formats:
1938                 for fmt in streaming_formats:
1939                     itag = str_or_none(fmt.get('itag'))
1940                     if not itag:
1941                         continue
1942                     quality = fmt.get('quality')
1943                     quality_label = fmt.get('qualityLabel') or quality
1944                     formats_spec[itag] = {
1945                         'asr': int_or_none(fmt.get('audioSampleRate')),
1946                         'filesize': int_or_none(fmt.get('contentLength')),
1947                         'format_note': quality_label,
1948                         'fps': int_or_none(fmt.get('fps')),
1949                         'height': int_or_none(fmt.get('height')),
1950                         'quality': q(quality),
1951                         # bitrate for itag 43 is always 2147483647
1952                         'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None,
1953                         'width': int_or_none(fmt.get('width')),
1954                     }
1955             formats = []
1956             for url_data_str in encoded_url_map.split(','):
1957                 url_data = compat_parse_qs(url_data_str)
1958                 if 'itag' not in url_data or 'url' not in url_data or url_data.get('drm_families'):
1959                     continue
1960                 stream_type = int_or_none(try_get(url_data, lambda x: x['stream_type'][0]))
1961                 # Unsupported FORMAT_STREAM_TYPE_OTF
1962                 if stream_type == 3:
1963                     continue
1964                 format_id = url_data['itag'][0]
1965                 url = url_data['url'][0]
1966
1967                 if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True):
1968                     ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
1969                     jsplayer_url_json = self._search_regex(
1970                         ASSETS_RE,
1971                         embed_webpage if age_gate else video_webpage,
1972                         'JS player URL (1)', default=None)
1973                     if not jsplayer_url_json and not age_gate:
1974                         # We need the embed website after all
1975                         if embed_webpage is None:
1976                             embed_url = proto + '://www.youtube.com/embed/%s' % video_id
1977                             embed_webpage = self._download_webpage(
1978                                 embed_url, video_id, 'Downloading embed webpage')
1979                         jsplayer_url_json = self._search_regex(
1980                             ASSETS_RE, embed_webpage, 'JS player URL')
1981
1982                     player_url = json.loads(jsplayer_url_json)
1983                     if player_url is None:
1984                         player_url_json = self._search_regex(
1985                             r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
1986                             video_webpage, 'age gate player URL')
1987                         player_url = json.loads(player_url_json)
1988
1989                 if 'sig' in url_data:
1990                     url += '&signature=' + url_data['sig'][0]
1991                 elif 's' in url_data:
1992                     encrypted_sig = url_data['s'][0]
1993
1994                     if self._downloader.params.get('verbose'):
1995                         if player_url is None:
1996                             player_version = 'unknown'
1997                             player_desc = 'unknown'
1998                         else:
1999                             if player_url.endswith('swf'):
2000                                 player_version = self._search_regex(
2001                                     r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
2002                                     'flash player', fatal=False)
2003                                 player_desc = 'flash player %s' % player_version
2004                             else:
2005                                 player_version = self._search_regex(
2006                                     [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
2007                                      r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'],
2008                                     player_url,
2009                                     'html5 player', fatal=False)
2010                                 player_desc = 'html5 player %s' % player_version
2011
2012                         parts_sizes = self._signature_cache_id(encrypted_sig)
2013                         self.to_screen('{%s} signature length %s, %s' %
2014                                        (format_id, parts_sizes, player_desc))
2015
2016                     signature = self._decrypt_signature(
2017                         encrypted_sig, video_id, player_url, age_gate)
2018                     sp = try_get(url_data, lambda x: x['sp'][0], compat_str) or 'signature'
2019                     url += '&%s=%s' % (sp, signature)
2020                 if 'ratebypass' not in url:
2021                     url += '&ratebypass=yes'
2022
2023                 dct = {
2024                     'format_id': format_id,
2025                     'url': url,
2026                     'player_url': player_url,
2027                 }
2028                 if format_id in self._formats:
2029                     dct.update(self._formats[format_id])
2030                 if format_id in formats_spec:
2031                     dct.update(formats_spec[format_id])
2032
2033                 # Some itags are not included in DASH manifest thus corresponding formats will
2034                 # lack metadata (see https://github.com/ytdl-org/youtube-dl/pull/5993).
2035                 # Trying to extract metadata from url_encoded_fmt_stream_map entry.
2036                 mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0])
2037                 width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None)
2038
2039                 filesize = int_or_none(url_data.get(
2040                     'clen', [None])[0]) or _extract_filesize(url)
2041
2042                 quality = url_data.get('quality', [None])[0]
2043
2044                 more_fields = {
2045                     'filesize': filesize,
2046                     'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000),
2047                     'width': width,
2048                     'height': height,
2049                     'fps': int_or_none(url_data.get('fps', [None])[0]),
2050                     'format_note': url_data.get('quality_label', [None])[0] or quality,
2051                     'quality': q(quality),
2052                 }
2053                 for key, value in more_fields.items():
2054                     if value:
2055                         dct[key] = value
2056                 type_ = url_data.get('type', [None])[0]
2057                 if type_:
2058                     type_split = type_.split(';')
2059                     kind_ext = type_split[0].split('/')
2060                     if len(kind_ext) == 2:
2061                         kind, _ = kind_ext
2062                         dct['ext'] = mimetype2ext(type_split[0])
2063                         if kind in ('audio', 'video'):
2064                             codecs = None
2065                             for mobj in re.finditer(
2066                                     r'(?P<key>[a-zA-Z_-]+)=(?P<quote>["\']?)(?P<val>.+?)(?P=quote)(?:;|$)', type_):
2067                                 if mobj.group('key') == 'codecs':
2068                                     codecs = mobj.group('val')
2069                                     break
2070                             if codecs:
2071                                 dct.update(parse_codecs(codecs))
2072                 if dct.get('acodec') == 'none' or dct.get('vcodec') == 'none':
2073                     dct['downloader_options'] = {
2074                         # Youtube throttles chunks >~10M
2075                         'http_chunk_size': 10485760,
2076                     }
2077                 formats.append(dct)
2078         else:
2079             manifest_url = (
2080                 url_or_none(try_get(
2081                     player_response,
2082                     lambda x: x['streamingData']['hlsManifestUrl'],
2083                     compat_str))
2084                 or url_or_none(try_get(
2085                     video_info, lambda x: x['hlsvp'][0], compat_str)))
2086             if manifest_url:
2087                 formats = []
2088                 m3u8_formats = self._extract_m3u8_formats(
2089                     manifest_url, video_id, 'mp4', fatal=False)
2090                 for a_format in m3u8_formats:
2091                     itag = self._search_regex(
2092                         r'/itag/(\d+)/', a_format['url'], 'itag', default=None)
2093                     if itag:
2094                         a_format['format_id'] = itag
2095                         if itag in self._formats:
2096                             dct = self._formats[itag].copy()
2097                             dct.update(a_format)
2098                             a_format = dct
2099                     a_format['player_url'] = player_url
2100                     # Accept-Encoding header causes failures in live streams on Youtube and Youtube Gaming
2101                     a_format.setdefault('http_headers', {})['Youtubedl-no-compression'] = 'True'
2102                     formats.append(a_format)
2103             else:
2104                 error_message = extract_unavailable_message()
2105                 if not error_message:
2106                     error_message = clean_html(try_get(
2107                         player_response, lambda x: x['playabilityStatus']['reason'],
2108                         compat_str))
2109                 if not error_message:
2110                     error_message = clean_html(
2111                         try_get(video_info, lambda x: x['reason'][0], compat_str))
2112                 if error_message:
2113                     raise ExtractorError(error_message, expected=True)
2114                 raise ExtractorError('no conn, hlsvp, hlsManifestUrl or url_encoded_fmt_stream_map information found in video info')
2115
2116         # uploader
2117         video_uploader = try_get(
2118             video_info, lambda x: x['author'][0],
2119             compat_str) or str_or_none(video_details.get('author'))
2120         if video_uploader:
2121             video_uploader = compat_urllib_parse_unquote_plus(video_uploader)
2122         else:
2123             self._downloader.report_warning('unable to extract uploader name')
2124
2125         # uploader_id
2126         video_uploader_id = None
2127         video_uploader_url = None
2128         mobj = re.search(
2129             r'<link itemprop="url" href="(?P<uploader_url>https?://www\.youtube\.com/(?:user|channel)/(?P<uploader_id>[^"]+))">',
2130             video_webpage)
2131         if mobj is not None:
2132             video_uploader_id = mobj.group('uploader_id')
2133             video_uploader_url = mobj.group('uploader_url')
2134         else:
2135             self._downloader.report_warning('unable to extract uploader nickname')
2136
2137         channel_id = (
2138             str_or_none(video_details.get('channelId'))
2139             or self._html_search_meta(
2140                 'channelId', video_webpage, 'channel id', default=None)
2141             or self._search_regex(
2142                 r'data-channel-external-id=(["\'])(?P<id>(?:(?!\1).)+)\1',
2143                 video_webpage, 'channel id', default=None, group='id'))
2144         channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None
2145
2146         # thumbnail image
2147         # We try first to get a high quality image:
2148         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
2149                             video_webpage, re.DOTALL)
2150         if m_thumb is not None:
2151             video_thumbnail = m_thumb.group(1)
2152         elif 'thumbnail_url' not in video_info:
2153             self._downloader.report_warning('unable to extract video thumbnail')
2154             video_thumbnail = None
2155         else:   # don't panic if we can't find it
2156             video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0])
2157
2158         # upload date
2159         upload_date = self._html_search_meta(
2160             'datePublished', video_webpage, 'upload date', default=None)
2161         if not upload_date:
2162             upload_date = self._search_regex(
2163                 [r'(?s)id="eow-date.*?>(.*?)</span>',
2164                  r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],
2165                 video_webpage, 'upload date', default=None)
2166         upload_date = unified_strdate(upload_date)
2167
2168         video_license = self._html_search_regex(
2169             r'<h4[^>]+class="title"[^>]*>\s*License\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li',
2170             video_webpage, 'license', default=None)
2171
2172         m_music = re.search(
2173             r'''(?x)
2174                 <h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*
2175                 <ul[^>]*>\s*
2176                 <li>(?P<title>.+?)
2177                 by (?P<creator>.+?)
2178                 (?:
2179                     \(.+?\)|
2180                     <a[^>]*
2181                         (?:
2182                             \bhref=["\']/red[^>]*>|             # drop possible
2183                             >\s*Listen ad-free with YouTube Red # YouTube Red ad
2184                         )
2185                     .*?
2186                 )?</li
2187             ''',
2188             video_webpage)
2189         if m_music:
2190             video_alt_title = remove_quotes(unescapeHTML(m_music.group('title')))
2191             video_creator = clean_html(m_music.group('creator'))
2192         else:
2193             video_alt_title = video_creator = None
2194
2195         def extract_meta(field):
2196             return self._html_search_regex(
2197                 r'<h4[^>]+class="title"[^>]*>\s*%s\s*</h4>\s*<ul[^>]*>\s*<li>(.+?)</li>\s*' % field,
2198                 video_webpage, field, default=None)
2199
2200         track = extract_meta('Song')
2201         artist = extract_meta('Artist')
2202         album = extract_meta('Album')
2203
2204         # Youtube Music Auto-generated description
2205         release_date = release_year = None
2206         if video_description:
2207             mobj = re.search(r'(?s)Provided to YouTube by [^\n]+\n+(?P<track>[^·]+)·(?P<artist>[^\n]+)\n+(?P<album>[^\n]+)(?:.+?℗\s*(?P<release_year>\d{4})(?!\d))?(?:.+?Released on\s*:\s*(?P<release_date>\d{4}-\d{2}-\d{2}))?(.+?\nArtist\s*:\s*(?P<clean_artist>[^\n]+))?', video_description)
2208             if mobj:
2209                 if not track:
2210                     track = mobj.group('track').strip()
2211                 if not artist:
2212                     artist = mobj.group('clean_artist') or ', '.join(a.strip() for a in mobj.group('artist').split('·'))
2213                 if not album:
2214                     album = mobj.group('album'.strip())
2215                 release_year = mobj.group('release_year')
2216                 release_date = mobj.group('release_date')
2217                 if release_date:
2218                     release_date = release_date.replace('-', '')
2219                     if not release_year:
2220                         release_year = int(release_date[:4])
2221                 if release_year:
2222                     release_year = int(release_year)
2223
2224         m_episode = re.search(
2225             r'<div[^>]+id="watch7-headline"[^>]*>\s*<span[^>]*>.*?>(?P<series>[^<]+)</a></b>\s*S(?P<season>\d+)\s*•\s*E(?P<episode>\d+)</span>',
2226             video_webpage)
2227         if m_episode:
2228             series = unescapeHTML(m_episode.group('series'))
2229             season_number = int(m_episode.group('season'))
2230             episode_number = int(m_episode.group('episode'))
2231         else:
2232             series = season_number = episode_number = None
2233
2234         m_cat_container = self._search_regex(
2235             r'(?s)<h4[^>]*>\s*Category\s*</h4>\s*<ul[^>]*>(.*?)</ul>',
2236             video_webpage, 'categories', default=None)
2237         if m_cat_container:
2238             category = self._html_search_regex(
2239                 r'(?s)<a[^<]+>(.*?)</a>', m_cat_container, 'category',
2240                 default=None)
2241             video_categories = None if category is None else [category]
2242         else:
2243             video_categories = None
2244
2245         video_tags = [
2246             unescapeHTML(m.group('content'))
2247             for m in re.finditer(self._meta_regex('og:video:tag'), video_webpage)]
2248
2249         def _extract_count(count_name):
2250             return str_to_int(self._search_regex(
2251                 r'-%s-button[^>]+><span[^>]+class="yt-uix-button-content"[^>]*>([\d,]+)</span>'
2252                 % re.escape(count_name),
2253                 video_webpage, count_name, default=None))
2254
2255         like_count = _extract_count('like')
2256         dislike_count = _extract_count('dislike')
2257
2258         if view_count is None:
2259             view_count = str_to_int(self._search_regex(
2260                 r'<[^>]+class=["\']watch-view-count[^>]+>\s*([\d,\s]+)', video_webpage,
2261                 'view count', default=None))
2262
2263         average_rating = (
2264             float_or_none(video_details.get('averageRating'))
2265             or try_get(video_info, lambda x: float_or_none(x['avg_rating'][0])))
2266
2267         # subtitles
2268         video_subtitles = self.extract_subtitles(video_id, video_webpage)
2269         automatic_captions = self.extract_automatic_captions(video_id, video_webpage)
2270
2271         video_duration = try_get(
2272             video_info, lambda x: int_or_none(x['length_seconds'][0]))
2273         if not video_duration:
2274             video_duration = int_or_none(video_details.get('lengthSeconds'))
2275         if not video_duration:
2276             video_duration = parse_duration(self._html_search_meta(
2277                 'duration', video_webpage, 'video duration'))
2278
2279         # annotations
2280         video_annotations = None
2281         if self._downloader.params.get('writeannotations', False):
2282             xsrf_token = self._search_regex(
2283                 r'([\'"])XSRF_TOKEN\1\s*:\s*([\'"])(?P<xsrf_token>[A-Za-z0-9+/=]+)\2',
2284                 video_webpage, 'xsrf token', group='xsrf_token', fatal=False)
2285             invideo_url = try_get(
2286                 player_response, lambda x: x['annotations'][0]['playerAnnotationsUrlsRenderer']['invideoUrl'], compat_str)
2287             if xsrf_token and invideo_url:
2288                 xsrf_field_name = self._search_regex(
2289                     r'([\'"])XSRF_FIELD_NAME\1\s*:\s*([\'"])(?P<xsrf_field_name>\w+)\2',
2290                     video_webpage, 'xsrf field name',
2291                     group='xsrf_field_name', default='session_token')
2292                 video_annotations = self._download_webpage(
2293                     self._proto_relative_url(invideo_url),
2294                     video_id, note='Downloading annotations',
2295                     errnote='Unable to download video annotations', fatal=False,
2296                     data=urlencode_postdata({xsrf_field_name: xsrf_token}))
2297
2298         chapters = self._extract_chapters(description_original, video_duration)
2299
2300         # Look for the DASH manifest
2301         if self._downloader.params.get('youtube_include_dash_manifest', True):
2302             dash_mpd_fatal = True
2303             for mpd_url in dash_mpds:
2304                 dash_formats = {}
2305                 try:
2306                     def decrypt_sig(mobj):
2307                         s = mobj.group(1)
2308                         dec_s = self._decrypt_signature(s, video_id, player_url, age_gate)
2309                         return '/signature/%s' % dec_s
2310
2311                     mpd_url = re.sub(r'/s/([a-fA-F0-9\.]+)', decrypt_sig, mpd_url)
2312
2313                     for df in self._extract_mpd_formats(
2314                             mpd_url, video_id, fatal=dash_mpd_fatal,
2315                             formats_dict=self._formats):
2316                         if not df.get('filesize'):
2317                             df['filesize'] = _extract_filesize(df['url'])
2318                         # Do not overwrite DASH format found in some previous DASH manifest
2319                         if df['format_id'] not in dash_formats:
2320                             dash_formats[df['format_id']] = df
2321                         # Additional DASH manifests may end up in HTTP Error 403 therefore
2322                         # allow them to fail without bug report message if we already have
2323                         # some DASH manifest succeeded. This is temporary workaround to reduce
2324                         # burst of bug reports until we figure out the reason and whether it
2325                         # can be fixed at all.
2326                         dash_mpd_fatal = False
2327                 except (ExtractorError, KeyError) as e:
2328                     self.report_warning(
2329                         'Skipping DASH manifest: %r' % e, video_id)
2330                 if dash_formats:
2331                     # Remove the formats we found through non-DASH, they
2332                     # contain less info and it can be wrong, because we use
2333                     # fixed values (for example the resolution). See
2334                     # https://github.com/ytdl-org/youtube-dl/issues/5774 for an
2335                     # example.
2336                     formats = [f for f in formats if f['format_id'] not in dash_formats.keys()]
2337                     formats.extend(dash_formats.values())
2338
2339         # Check for malformed aspect ratio
2340         stretched_m = re.search(
2341             r'<meta\s+property="og:video:tag".*?content="yt:stretch=(?P<w>[0-9]+):(?P<h>[0-9]+)">',
2342             video_webpage)
2343         if stretched_m:
2344             w = float(stretched_m.group('w'))
2345             h = float(stretched_m.group('h'))
2346             # yt:stretch may hold invalid ratio data (e.g. for Q39EVAstoRM ratio is 17:0).
2347             # We will only process correct ratios.
2348             if w > 0 and h > 0:
2349                 ratio = w / h
2350                 for f in formats:
2351                     if f.get('vcodec') != 'none':
2352                         f['stretched_ratio'] = ratio
2353
2354         if not formats:
2355             token = extract_token(video_info)
2356             if not token:
2357                 if 'reason' in video_info:
2358                     if 'The uploader has not made this video available in your country.' in video_info['reason']:
2359                         regions_allowed = self._html_search_meta(
2360                             'regionsAllowed', video_webpage, default=None)
2361                         countries = regions_allowed.split(',') if regions_allowed else None
2362                         self.raise_geo_restricted(
2363                             msg=video_info['reason'][0], countries=countries)
2364                     reason = video_info['reason'][0]
2365                     if 'Invalid parameters' in reason:
2366                         unavailable_message = extract_unavailable_message()
2367                         if unavailable_message:
2368                             reason = unavailable_message
2369                     raise ExtractorError(
2370                         'YouTube said: %s' % reason,
2371                         expected=True, video_id=video_id)
2372                 else:
2373                     raise ExtractorError(
2374                         '"token" parameter not in video info for unknown reason',
2375                         video_id=video_id)
2376
2377         if not formats and (video_info.get('license_info') or try_get(player_response, lambda x: x['streamingData']['licenseInfos'])):
2378             raise ExtractorError('This video is DRM protected.', expected=True)
2379
2380         self._sort_formats(formats)
2381
2382         self.mark_watched(video_id, video_info, player_response)
2383
2384         return {
2385             'id': video_id,
2386             'uploader': video_uploader,
2387             'uploader_id': video_uploader_id,
2388             'uploader_url': video_uploader_url,
2389             'channel_id': channel_id,
2390             'channel_url': channel_url,
2391             'upload_date': upload_date,
2392             'license': video_license,
2393             'creator': video_creator or artist,
2394             'title': video_title,
2395             'alt_title': video_alt_title or track,
2396             'thumbnail': video_thumbnail,
2397             'description': video_description,
2398             'categories': video_categories,
2399             'tags': video_tags,
2400             'subtitles': video_subtitles,
2401             'automatic_captions': automatic_captions,
2402             'duration': video_duration,
2403             'age_limit': 18 if age_gate else 0,
2404             'annotations': video_annotations,
2405             'chapters': chapters,
2406             'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id,
2407             'view_count': view_count,
2408             'like_count': like_count,
2409             'dislike_count': dislike_count,
2410             'average_rating': average_rating,
2411             'formats': formats,
2412             'is_live': is_live,
2413             'start_time': start_time,
2414             'end_time': end_time,
2415             'series': series,
2416             'season_number': season_number,
2417             'episode_number': episode_number,
2418             'track': track,
2419             'artist': artist,
2420             'album': album,
2421             'release_date': release_date,
2422             'release_year': release_year,
2423         }
2424
2425
2426 class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
2427     IE_DESC = 'YouTube.com playlists'
2428     _VALID_URL = r"""(?x)(?:
2429                         (?:https?://)?
2430                         (?:\w+\.)?
2431                         (?:
2432                             (?:
2433                                 youtube\.com|
2434                                 invidio\.us
2435                             )
2436                             /
2437                             (?:
2438                                (?:course|view_play_list|my_playlists|artist|playlist|watch|embed/(?:videoseries|[0-9A-Za-z_-]{11}))
2439                                \? (?:.*?[&;])*? (?:p|a|list)=
2440                             |  p/
2441                             )|
2442                             youtu\.be/[0-9A-Za-z_-]{11}\?.*?\blist=
2443                         )
2444                         (
2445                             (?:PL|LL|EC|UU|FL|RD|UL|TL|OLAK5uy_)?[0-9A-Za-z-_]{10,}
2446                             # Top tracks, they can also include dots
2447                             |(?:MC)[\w\.]*
2448                         )
2449                         .*
2450                      |
2451                         (%(playlist_id)s)
2452                      )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
2453     _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
2454     _VIDEO_RE_TPL = r'href="\s*/watch\?v=%s(?:&amp;(?:[^"]*?index=(?P<index>\d+))?(?:[^>]+>(?P<title>[^<]+))?)?'
2455     _VIDEO_RE = _VIDEO_RE_TPL % r'(?P<id>[0-9A-Za-z_-]{11})'
2456     IE_NAME = 'youtube:playlist'
2457     _TESTS = [{
2458         'url': 'https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
2459         'info_dict': {
2460             'title': 'ytdl test PL',
2461             'id': 'PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re',
2462         },
2463         'playlist_count': 3,
2464     }, {
2465         'url': 'https://www.youtube.com/playlist?list=PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
2466         'info_dict': {
2467             'id': 'PLtPgu7CB4gbZDA7i_euNxn75ISqxwZPYx',
2468             'title': 'YDL_Empty_List',
2469         },
2470         'playlist_count': 0,
2471         'skip': 'This playlist is private',
2472     }, {
2473         'note': 'Playlist with deleted videos (#651). As a bonus, the video #51 is also twice in this list.',
2474         'url': 'https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
2475         'info_dict': {
2476             'title': '29C3: Not my department',
2477             'id': 'PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC',
2478             'uploader': 'Christiaan008',
2479             'uploader_id': 'ChRiStIaAn008',
2480         },
2481         'playlist_count': 95,
2482     }, {
2483         'note': 'issue #673',
2484         'url': 'PLBB231211A4F62143',
2485         'info_dict': {
2486             'title': '[OLD]Team Fortress 2 (Class-based LP)',
2487             'id': 'PLBB231211A4F62143',
2488             'uploader': 'Wickydoo',
2489             'uploader_id': 'Wickydoo',
2490         },
2491         'playlist_mincount': 26,
2492     }, {
2493         'note': 'Large playlist',
2494         'url': 'https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q',
2495         'info_dict': {
2496             'title': 'Uploads from Cauchemar',
2497             'id': 'UUBABnxM4Ar9ten8Mdjj1j0Q',
2498             'uploader': 'Cauchemar',
2499             'uploader_id': 'Cauchemar89',
2500         },
2501         'playlist_mincount': 799,
2502     }, {
2503         'url': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
2504         'info_dict': {
2505             'title': 'YDL_safe_search',
2506             'id': 'PLtPgu7CB4gbY9oDN3drwC3cMbJggS7dKl',
2507         },
2508         'playlist_count': 2,
2509         'skip': 'This playlist is private',
2510     }, {
2511         'note': 'embedded',
2512         'url': 'https://www.youtube.com/embed/videoseries?list=PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
2513         'playlist_count': 4,
2514         'info_dict': {
2515             'title': 'JODA15',
2516             'id': 'PL6IaIsEjSbf96XFRuNccS_RuEXwNdsoEu',
2517             'uploader': 'milan',
2518             'uploader_id': 'UCEI1-PVPcYXjB73Hfelbmaw',
2519         }
2520     }, {
2521         'url': 'http://www.youtube.com/embed/_xDOZElKyNU?list=PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
2522         'playlist_mincount': 485,
2523         'info_dict': {
2524             'title': '2018 Chinese New Singles (11/6 updated)',
2525             'id': 'PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
2526             'uploader': 'LBK',
2527             'uploader_id': 'sdragonfang',
2528         }
2529     }, {
2530         'note': 'Embedded SWF player',
2531         'url': 'https://www.youtube.com/p/YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ?hl=en_US&fs=1&rel=0',
2532         'playlist_count': 4,
2533         'info_dict': {
2534             'title': 'JODA7',
2535             'id': 'YN5VISEtHet5D4NEvfTd0zcgFk84NqFZ',
2536         },
2537         'skip': 'This playlist does not exist',
2538     }, {
2539         'note': 'Buggy playlist: the webpage has a "Load more" button but it doesn\'t have more videos',
2540         'url': 'https://www.youtube.com/playlist?list=UUXw-G3eDE9trcvY2sBMM_aA',
2541         'info_dict': {
2542             'title': 'Uploads from Interstellar Movie',
2543             'id': 'UUXw-G3eDE9trcvY2sBMM_aA',
2544             'uploader': 'Interstellar Movie',
2545             'uploader_id': 'InterstellarMovie1',
2546         },
2547         'playlist_mincount': 21,
2548     }, {
2549         # Playlist URL that does not actually serve a playlist
2550         'url': 'https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4',
2551         'info_dict': {
2552             'id': 'FqZTN594JQw',
2553             'ext': 'webm',
2554             'title': "Smiley's People 01 detective, Adventure Series, Action",
2555             'uploader': 'STREEM',
2556             'uploader_id': 'UCyPhqAZgwYWZfxElWVbVJng',
2557             'uploader_url': r're:https?://(?:www\.)?youtube\.com/channel/UCyPhqAZgwYWZfxElWVbVJng',
2558             'upload_date': '20150526',
2559             'license': 'Standard YouTube License',
2560             'description': 'md5:507cdcb5a49ac0da37a920ece610be80',
2561             'categories': ['People & Blogs'],
2562             'tags': list,
2563             'view_count': int,
2564             'like_count': int,
2565             'dislike_count': int,
2566         },
2567         'params': {
2568             'skip_download': True,
2569         },
2570         'skip': 'This video is not available.',
2571         'add_ie': [YoutubeIE.ie_key()],
2572     }, {
2573         'url': 'https://youtu.be/yeWKywCrFtk?list=PL2qgrgXsNUG5ig9cat4ohreBjYLAPC0J5',
2574         'info_dict': {
2575             'id': 'yeWKywCrFtk',
2576             'ext': 'mp4',
2577             'title': 'Small Scale Baler and Braiding Rugs',
2578             'uploader': 'Backus-Page House Museum',
2579             'uploader_id': 'backuspagemuseum',
2580             'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/backuspagemuseum',
2581             'upload_date': '20161008',
2582             'description': 'md5:800c0c78d5eb128500bffd4f0b4f2e8a',
2583             'categories': ['Nonprofits & Activism'],
2584             'tags': list,
2585             'like_count': int,
2586             'dislike_count': int,
2587         },
2588         'params': {
2589             'noplaylist': True,
2590             'skip_download': True,
2591         },
2592     }, {
2593         # https://github.com/ytdl-org/youtube-dl/issues/21844
2594         'url': 'https://www.youtube.com/playlist?list=PLzH6n4zXuckpfMu_4Ff8E7Z1behQks5ba',
2595         'info_dict': {
2596             'title': 'Data Analysis with Dr Mike Pound',
2597             'id': 'PLzH6n4zXuckpfMu_4Ff8E7Z1behQks5ba',
2598             'uploader_id': 'Computerphile',
2599             'uploader': 'Computerphile',
2600         },
2601         'playlist_mincount': 11,
2602     }, {
2603         'url': 'https://youtu.be/uWyaPkt-VOI?list=PL9D9FC436B881BA21',
2604         'only_matching': True,
2605     }, {
2606         'url': 'TLGGrESM50VT6acwMjAyMjAxNw',
2607         'only_matching': True,
2608     }, {
2609         # music album playlist
2610         'url': 'OLAK5uy_m4xAFdmMC5rX3Ji3g93pQe3hqLZw_9LhM',
2611         'only_matching': True,
2612     }, {
2613         'url': 'https://invidio.us/playlist?list=PLDIoUOhQQPlXr63I_vwF9GD8sAKh77dWU',
2614         'only_matching': True,
2615     }]
2616
2617     def _real_initialize(self):
2618         self._login()
2619
2620     def extract_videos_from_page(self, page):
2621         ids_in_page = []
2622         titles_in_page = []
2623
2624         for item in re.findall(
2625                 r'(<[^>]*\bdata-video-id\s*=\s*["\'][0-9A-Za-z_-]{11}[^>]+>)', page):
2626             attrs = extract_attributes(item)
2627             video_id = attrs['data-video-id']
2628             video_title = unescapeHTML(attrs.get('data-title'))
2629             if video_title:
2630                 video_title = video_title.strip()
2631             ids_in_page.append(video_id)
2632             titles_in_page.append(video_title)
2633
2634         # Fallback with old _VIDEO_RE
2635         self.extract_videos_from_page_impl(
2636             self._VIDEO_RE, page, ids_in_page, titles_in_page)
2637
2638         # Relaxed fallbacks
2639         self.extract_videos_from_page_impl(
2640             r'href="\s*/watch\?v\s*=\s*(?P<id>[0-9A-Za-z_-]{11})', page,
2641             ids_in_page, titles_in_page)
2642         self.extract_videos_from_page_impl(
2643             r'data-video-ids\s*=\s*["\'](?P<id>[0-9A-Za-z_-]{11})', page,
2644             ids_in_page, titles_in_page)
2645
2646         return zip(ids_in_page, titles_in_page)
2647
2648     def _extract_mix(self, playlist_id):
2649         # The mixes are generated from a single video
2650         # the id of the playlist is just 'RD' + video_id
2651         ids = []
2652         last_id = playlist_id[-11:]
2653         for n in itertools.count(1):
2654             url = 'https://youtube.com/watch?v=%s&list=%s' % (last_id, playlist_id)
2655             webpage = self._download_webpage(
2656                 url, playlist_id, 'Downloading page {0} of Youtube mix'.format(n))
2657             new_ids = orderedSet(re.findall(
2658                 r'''(?xs)data-video-username=".*?".*?
2659                            href="/watch\?v=([0-9A-Za-z_-]{11})&amp;[^"]*?list=%s''' % re.escape(playlist_id),
2660                 webpage))
2661             # Fetch new pages until all the videos are repeated, it seems that
2662             # there are always 51 unique videos.
2663             new_ids = [_id for _id in new_ids if _id not in ids]
2664             if not new_ids:
2665                 break
2666             ids.extend(new_ids)
2667             last_id = ids[-1]
2668
2669         url_results = self._ids_to_results(ids)
2670
2671         search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage)
2672         title_span = (
2673             search_title('playlist-title')
2674             or search_title('title long-title')
2675             or search_title('title'))
2676         title = clean_html(title_span)
2677
2678         return self.playlist_result(url_results, playlist_id, title)
2679
2680     def _extract_playlist(self, playlist_id):
2681         url = self._TEMPLATE_URL % playlist_id
2682         page = self._download_webpage(url, playlist_id)
2683
2684         # the yt-alert-message now has tabindex attribute (see https://github.com/ytdl-org/youtube-dl/issues/11604)
2685         for match in re.findall(r'<div class="yt-alert-message"[^>]*>([^<]+)</div>', page):
2686             match = match.strip()
2687             # Check if the playlist exists or is private
2688             mobj = re.match(r'[^<]*(?:The|This) playlist (?P<reason>does not exist|is private)[^<]*', match)
2689             if mobj:
2690                 reason = mobj.group('reason')
2691                 message = 'This playlist %s' % reason
2692                 if 'private' in reason:
2693                     message += ', use --username or --netrc to access it'
2694                 message += '.'
2695                 raise ExtractorError(message, expected=True)
2696             elif re.match(r'[^<]*Invalid parameters[^<]*', match):
2697                 raise ExtractorError(
2698                     'Invalid parameters. Maybe URL is incorrect.',
2699                     expected=True)
2700             elif re.match(r'[^<]*Choose your language[^<]*', match):
2701                 continue
2702             else:
2703                 self.report_warning('Youtube gives an alert message: ' + match)
2704
2705         playlist_title = self._html_search_regex(
2706             r'(?s)<h1 class="pl-header-title[^"]*"[^>]*>\s*(.*?)\s*</h1>',
2707             page, 'title', default=None)
2708
2709         _UPLOADER_BASE = r'class=["\']pl-header-details[^>]+>\s*<li>\s*<a[^>]+\bhref='
2710         uploader = self._search_regex(
2711             r'%s["\']/(?:user|channel)/[^>]+>([^<]+)' % _UPLOADER_BASE,
2712             page, 'uploader', default=None)
2713         mobj = re.search(
2714             r'%s(["\'])(?P<path>/(?:user|channel)/(?P<uploader_id>.+?))\1' % _UPLOADER_BASE,
2715             page)
2716         if mobj:
2717             uploader_id = mobj.group('uploader_id')
2718             uploader_url = compat_urlparse.urljoin(url, mobj.group('path'))
2719         else:
2720             uploader_id = uploader_url = None
2721
2722         has_videos = True
2723
2724         if not playlist_title:
2725             try:
2726                 # Some playlist URLs don't actually serve a playlist (e.g.
2727                 # https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4)
2728                 next(self._entries(page, playlist_id))
2729             except StopIteration:
2730                 has_videos = False
2731
2732         playlist = self.playlist_result(
2733             self._entries(page, playlist_id), playlist_id, playlist_title)
2734         playlist.update({
2735             'uploader': uploader,
2736             'uploader_id': uploader_id,
2737             'uploader_url': uploader_url,
2738         })
2739
2740         return has_videos, playlist
2741
2742     def _check_download_just_video(self, url, playlist_id):
2743         # Check if it's a video-specific URL
2744         query_dict = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
2745         video_id = query_dict.get('v', [None])[0] or self._search_regex(
2746             r'(?:(?:^|//)youtu\.be/|youtube\.com/embed/(?!videoseries))([0-9A-Za-z_-]{11})', url,
2747             'video id', default=None)
2748         if video_id:
2749             if self._downloader.params.get('noplaylist'):
2750                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
2751                 return video_id, self.url_result(video_id, 'Youtube', video_id=video_id)
2752             else:
2753                 self.to_screen('Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id))
2754                 return video_id, None
2755         return None, None
2756
2757     def _real_extract(self, url):
2758         # Extract playlist id
2759         mobj = re.match(self._VALID_URL, url)
2760         if mobj is None:
2761             raise ExtractorError('Invalid URL: %s' % url)
2762         playlist_id = mobj.group(1) or mobj.group(2)
2763
2764         video_id, video = self._check_download_just_video(url, playlist_id)
2765         if video:
2766             return video
2767
2768         if playlist_id.startswith(('RD', 'UL', 'PU')):
2769             # Mixes require a custom extraction process
2770             return self._extract_mix(playlist_id)
2771
2772         has_videos, playlist = self._extract_playlist(playlist_id)
2773         if has_videos or not video_id:
2774             return playlist
2775
2776         # Some playlist URLs don't actually serve a playlist (see
2777         # https://github.com/ytdl-org/youtube-dl/issues/10537).
2778         # Fallback to plain video extraction if there is a video id
2779         # along with playlist id.
2780         return self.url_result(video_id, 'Youtube', video_id=video_id)
2781
2782
2783 class YoutubeChannelIE(YoutubePlaylistBaseInfoExtractor):
2784     IE_DESC = 'YouTube.com channels'
2785     _VALID_URL = r'https?://(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com|(?:www\.)?invidio\.us)/channel/(?P<id>[0-9A-Za-z_-]+)'
2786     _TEMPLATE_URL = 'https://www.youtube.com/channel/%s/videos'
2787     _VIDEO_RE = r'(?:title="(?P<title>[^"]+)"[^>]+)?href="/watch\?v=(?P<id>[0-9A-Za-z_-]+)&?'
2788     IE_NAME = 'youtube:channel'
2789     _TESTS = [{
2790         'note': 'paginated channel',
2791         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w',
2792         'playlist_mincount': 91,
2793         'info_dict': {
2794             'id': 'UUKfVa3S1e4PHvxWcwyMMg8w',
2795             'title': 'Uploads from lex will',
2796             'uploader': 'lex will',
2797             'uploader_id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
2798         }
2799     }, {
2800         'note': 'Age restricted channel',
2801         # from https://www.youtube.com/user/DeusExOfficial
2802         'url': 'https://www.youtube.com/channel/UCs0ifCMCm1icqRbqhUINa0w',
2803         'playlist_mincount': 64,
2804         'info_dict': {
2805             'id': 'UUs0ifCMCm1icqRbqhUINa0w',
2806             'title': 'Uploads from Deus Ex',
2807             'uploader': 'Deus Ex',
2808             'uploader_id': 'DeusExOfficial',
2809         },
2810     }, {
2811         'url': 'https://invidio.us/channel/UC23qupoDRn9YOAVzeoxjOQA',
2812         'only_matching': True,
2813     }]
2814
2815     @classmethod
2816     def suitable(cls, url):
2817         return (False if YoutubePlaylistsIE.suitable(url) or YoutubeLiveIE.suitable(url)
2818                 else super(YoutubeChannelIE, cls).suitable(url))
2819
2820     def _build_template_url(self, url, channel_id):
2821         return self._TEMPLATE_URL % channel_id
2822
2823     def _real_extract(self, url):
2824         channel_id = self._match_id(url)
2825
2826         url = self._build_template_url(url, channel_id)
2827
2828         # Channel by page listing is restricted to 35 pages of 30 items, i.e. 1050 videos total (see #5778)
2829         # Workaround by extracting as a playlist if managed to obtain channel playlist URL
2830         # otherwise fallback on channel by page extraction
2831         channel_page = self._download_webpage(
2832             url + '?view=57', channel_id,
2833             'Downloading channel page', fatal=False)
2834         if channel_page is False:
2835             channel_playlist_id = False
2836         else:
2837             channel_playlist_id = self._html_search_meta(
2838                 'channelId', channel_page, 'channel id', default=None)
2839             if not channel_playlist_id:
2840                 channel_url = self._html_search_meta(
2841                     ('al:ios:url', 'twitter:app:url:iphone', 'twitter:app:url:ipad'),
2842                     channel_page, 'channel url', default=None)
2843                 if channel_url:
2844                     channel_playlist_id = self._search_regex(
2845                         r'vnd\.youtube://user/([0-9A-Za-z_-]+)',
2846                         channel_url, 'channel id', default=None)
2847         if channel_playlist_id and channel_playlist_id.startswith('UC'):
2848             playlist_id = 'UU' + channel_playlist_id[2:]
2849             return self.url_result(
2850                 compat_urlparse.urljoin(url, '/playlist?list=%s' % playlist_id), 'YoutubePlaylist')
2851
2852         channel_page = self._download_webpage(url, channel_id, 'Downloading page #1')
2853         autogenerated = re.search(r'''(?x)
2854                 class="[^"]*?(?:
2855                     channel-header-autogenerated-label|
2856                     yt-channel-title-autogenerated
2857                 )[^"]*"''', channel_page) is not None
2858
2859         if autogenerated:
2860             # The videos are contained in a single page
2861             # the ajax pages can't be used, they are empty
2862             entries = [
2863                 self.url_result(
2864                     video_id, 'Youtube', video_id=video_id,
2865                     video_title=video_title)
2866                 for video_id, video_title in self.extract_videos_from_page(channel_page)]
2867             return self.playlist_result(entries, channel_id)
2868
2869         try:
2870             next(self._entries(channel_page, channel_id))
2871         except StopIteration:
2872             alert_message = self._html_search_regex(
2873                 r'(?s)<div[^>]+class=(["\']).*?\byt-alert-message\b.*?\1[^>]*>(?P<alert>[^<]+)</div>',
2874                 channel_page, 'alert', default=None, group='alert')
2875             if alert_message:
2876                 raise ExtractorError('Youtube said: %s' % alert_message, expected=True)
2877
2878         return self.playlist_result(self._entries(channel_page, channel_id), channel_id)
2879
2880
2881 class YoutubeUserIE(YoutubeChannelIE):
2882     IE_DESC = 'YouTube.com user videos (URL or "ytuser" keyword)'
2883     _VALID_URL = r'(?:(?:https?://(?:\w+\.)?youtube\.com/(?:(?P<user>user|c)/)?(?!(?:attribution_link|watch|results|shared)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)(?P<id>[A-Za-z0-9_-]+)'
2884     _TEMPLATE_URL = 'https://www.youtube.com/%s/%s/videos'
2885     IE_NAME = 'youtube:user'
2886
2887     _TESTS = [{
2888         'url': 'https://www.youtube.com/user/TheLinuxFoundation',
2889         'playlist_mincount': 320,
2890         'info_dict': {
2891             'id': 'UUfX55Sx5hEFjoC3cNs6mCUQ',
2892             'title': 'Uploads from The Linux Foundation',
2893             'uploader': 'The Linux Foundation',
2894             'uploader_id': 'TheLinuxFoundation',
2895         }
2896     }, {
2897         # Only available via https://www.youtube.com/c/12minuteathlete/videos
2898         # but not https://www.youtube.com/user/12minuteathlete/videos
2899         'url': 'https://www.youtube.com/c/12minuteathlete/videos',
2900         'playlist_mincount': 249,
2901         'info_dict': {
2902             'id': 'UUVjM-zV6_opMDx7WYxnjZiQ',
2903             'title': 'Uploads from 12 Minute Athlete',
2904             'uploader': '12 Minute Athlete',
2905             'uploader_id': 'the12minuteathlete',
2906         }
2907     }, {
2908         'url': 'ytuser:phihag',
2909         'only_matching': True,
2910     }, {
2911         'url': 'https://www.youtube.com/c/gametrailers',
2912         'only_matching': True,
2913     }, {
2914         'url': 'https://www.youtube.com/gametrailers',
2915         'only_matching': True,
2916     }, {
2917         # This channel is not available, geo restricted to JP
2918         'url': 'https://www.youtube.com/user/kananishinoSMEJ/videos',
2919         'only_matching': True,
2920     }]
2921
2922     @classmethod
2923     def suitable(cls, url):
2924         # Don't return True if the url can be extracted with other youtube
2925         # extractor, the regex would is too permissive and it would match.
2926         other_yt_ies = iter(klass for (name, klass) in globals().items() if name.startswith('Youtube') and name.endswith('IE') and klass is not cls)
2927         if any(ie.suitable(url) for ie in other_yt_ies):
2928             return False
2929         else:
2930             return super(YoutubeUserIE, cls).suitable(url)
2931
2932     def _build_template_url(self, url, channel_id):
2933         mobj = re.match(self._VALID_URL, url)
2934         return self._TEMPLATE_URL % (mobj.group('user') or 'user', mobj.group('id'))
2935
2936
2937 class YoutubeLiveIE(YoutubeBaseInfoExtractor):
2938     IE_DESC = 'YouTube.com live streams'
2939     _VALID_URL = r'(?P<base_url>https?://(?:\w+\.)?youtube\.com/(?:(?:user|channel|c)/)?(?P<id>[^/]+))/live'
2940     IE_NAME = 'youtube:live'
2941
2942     _TESTS = [{
2943         'url': 'https://www.youtube.com/user/TheYoungTurks/live',
2944         'info_dict': {
2945             'id': 'a48o2S1cPoo',
2946             'ext': 'mp4',
2947             'title': 'The Young Turks - Live Main Show',
2948             'uploader': 'The Young Turks',
2949             'uploader_id': 'TheYoungTurks',
2950             'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/TheYoungTurks',
2951             'upload_date': '20150715',
2952             'license': 'Standard YouTube License',
2953             'description': 'md5:438179573adcdff3c97ebb1ee632b891',
2954             'categories': ['News & Politics'],
2955             'tags': ['Cenk Uygur (TV Program Creator)', 'The Young Turks (Award-Winning Work)', 'Talk Show (TV Genre)'],
2956             'like_count': int,
2957             'dislike_count': int,
2958         },
2959         'params': {
2960             'skip_download': True,
2961         },
2962     }, {
2963         'url': 'https://www.youtube.com/channel/UC1yBKRuGpC1tSM73A0ZjYjQ/live',
2964         'only_matching': True,
2965     }, {
2966         'url': 'https://www.youtube.com/c/CommanderVideoHq/live',
2967         'only_matching': True,
2968     }, {
2969         'url': 'https://www.youtube.com/TheYoungTurks/live',
2970         'only_matching': True,
2971     }]
2972
2973     def _real_extract(self, url):
2974         mobj = re.match(self._VALID_URL, url)
2975         channel_id = mobj.group('id')
2976         base_url = mobj.group('base_url')
2977         webpage = self._download_webpage(url, channel_id, fatal=False)
2978         if webpage:
2979             page_type = self._og_search_property(
2980                 'type', webpage, 'page type', default='')
2981             video_id = self._html_search_meta(
2982                 'videoId', webpage, 'video id', default=None)
2983             if page_type.startswith('video') and video_id and re.match(
2984                     r'^[0-9A-Za-z_-]{11}$', video_id):
2985                 return self.url_result(video_id, YoutubeIE.ie_key())
2986         return self.url_result(base_url)
2987
2988
2989 class YoutubePlaylistsIE(YoutubePlaylistsBaseInfoExtractor):
2990     IE_DESC = 'YouTube.com user/channel playlists'
2991     _VALID_URL = r'https?://(?:\w+\.)?youtube\.com/(?:user|channel)/(?P<id>[^/]+)/playlists'
2992     IE_NAME = 'youtube:playlists'
2993
2994     _TESTS = [{
2995         'url': 'https://www.youtube.com/user/ThirstForScience/playlists',
2996         'playlist_mincount': 4,
2997         'info_dict': {
2998             'id': 'ThirstForScience',
2999             'title': 'ThirstForScience',
3000         },
3001     }, {
3002         # with "Load more" button
3003         'url': 'https://www.youtube.com/user/igorkle1/playlists?view=1&sort=dd',
3004         'playlist_mincount': 70,
3005         'info_dict': {
3006             'id': 'igorkle1',
3007             'title': 'Игорь Клейнер',
3008         },
3009     }, {
3010         'url': 'https://www.youtube.com/channel/UCiU1dHvZObB2iP6xkJ__Icw/playlists',
3011         'playlist_mincount': 17,
3012         'info_dict': {
3013             'id': 'UCiU1dHvZObB2iP6xkJ__Icw',
3014             'title': 'Chem Player',
3015         },
3016         'skip': 'Blocked',
3017     }]
3018
3019
3020 class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor):
3021     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
3022
3023
3024 class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor):
3025     IE_DESC = 'YouTube.com searches'
3026     # there doesn't appear to be a real limit, for example if you search for
3027     # 'python' you get more than 8.000.000 results
3028     _MAX_RESULTS = float('inf')
3029     IE_NAME = 'youtube:search'
3030     _SEARCH_KEY = 'ytsearch'
3031     _EXTRA_QUERY_ARGS = {}
3032     _TESTS = []
3033
3034     def _get_n_results(self, query, n):
3035         """Get a specified number of results for a query"""
3036
3037         videos = []
3038         limit = n
3039
3040         url_query = {
3041             'search_query': query.encode('utf-8'),
3042         }
3043         url_query.update(self._EXTRA_QUERY_ARGS)
3044         result_url = 'https://www.youtube.com/results?' + compat_urllib_parse_urlencode(url_query)
3045
3046         for pagenum in itertools.count(1):
3047             data = self._download_json(
3048                 result_url, video_id='query "%s"' % query,
3049                 note='Downloading page %s' % pagenum,
3050                 errnote='Unable to download API page',
3051                 query={'spf': 'navigate'})
3052             html_content = data[1]['body']['content']
3053
3054             if 'class="search-message' in html_content:
3055                 raise ExtractorError(
3056                     '[youtube] No video results', expected=True)
3057
3058             new_videos = list(self._process_page(html_content))
3059             videos += new_videos
3060             if not new_videos or len(videos) > limit:
3061                 break
3062             next_link = self._html_search_regex(
3063                 r'href="(/results\?[^"]*\bsp=[^"]+)"[^>]*>\s*<span[^>]+class="[^"]*\byt-uix-button-content\b[^"]*"[^>]*>Next',
3064                 html_content, 'next link', default=None)
3065             if next_link is None:
3066                 break
3067             result_url = compat_urlparse.urljoin('https://www.youtube.com/', next_link)
3068
3069         if len(videos) > n:
3070             videos = videos[:n]
3071         return self.playlist_result(videos, query)
3072
3073
3074 class YoutubeSearchDateIE(YoutubeSearchIE):
3075     IE_NAME = YoutubeSearchIE.IE_NAME + ':date'
3076     _SEARCH_KEY = 'ytsearchdate'
3077     IE_DESC = 'YouTube.com searches, newest videos first'
3078     _EXTRA_QUERY_ARGS = {'search_sort': 'video_date_uploaded'}
3079
3080
3081 class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
3082     IE_DESC = 'YouTube.com search URLs'
3083     IE_NAME = 'youtube:search_url'
3084     _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)'
3085     _TESTS = [{
3086         'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
3087         'playlist_mincount': 5,
3088         'info_dict': {
3089             'title': 'youtube-dl test video',
3090         }
3091     }, {
3092         'url': 'https://www.youtube.com/results?q=test&sp=EgQIBBgB',
3093         'only_matching': True,
3094     }]
3095
3096     def _real_extract(self, url):
3097         mobj = re.match(self._VALID_URL, url)
3098         query = compat_urllib_parse_unquote_plus(mobj.group('query'))
3099         webpage = self._download_webpage(url, query)
3100         return self.playlist_result(self._process_page(webpage), playlist_title=query)
3101
3102
3103 class YoutubeShowIE(YoutubePlaylistsBaseInfoExtractor):
3104     IE_DESC = 'YouTube.com (multi-season) shows'
3105     _VALID_URL = r'https?://(?:www\.)?youtube\.com/show/(?P<id>[^?#]*)'
3106     IE_NAME = 'youtube:show'
3107     _TESTS = [{
3108         'url': 'https://www.youtube.com/show/airdisasters',
3109         'playlist_mincount': 5,
3110         'info_dict': {
3111             'id': 'airdisasters',
3112             'title': 'Air Disasters',
3113         }
3114     }]
3115
3116     def _real_extract(self, url):
3117         playlist_id = self._match_id(url)
3118         return super(YoutubeShowIE, self)._real_extract(
3119             'https://www.youtube.com/show/%s/playlists' % playlist_id)
3120
3121
3122 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
3123     """
3124     Base class for feed extractors
3125     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
3126     """
3127     _LOGIN_REQUIRED = True
3128
3129     @property
3130     def IE_NAME(self):
3131         return 'youtube:%s' % self._FEED_NAME
3132
3133     def _real_initialize(self):
3134         self._login()
3135
3136     def _entries(self, page):
3137         # The extraction process is the same as for playlists, but the regex
3138         # for the video ids doesn't contain an index
3139         ids = []
3140         more_widget_html = content_html = page
3141         for page_num in itertools.count(1):
3142             matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html)
3143
3144             # 'recommended' feed has infinite 'load more' and each new portion spins
3145             # the same videos in (sometimes) slightly different order, so we'll check
3146             # for unicity and break when portion has no new videos
3147             new_ids = list(filter(lambda video_id: video_id not in ids, orderedSet(matches)))
3148             if not new_ids:
3149                 break
3150
3151             ids.extend(new_ids)
3152
3153             for entry in self._ids_to_results(new_ids):
3154                 yield entry
3155
3156             mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
3157             if not mobj:
3158                 break
3159
3160             more = self._download_json(
3161                 'https://youtube.com/%s' % mobj.group('more'), self._PLAYLIST_TITLE,
3162                 'Downloading page #%s' % page_num,
3163                 transform_source=uppercase_escape)
3164             content_html = more['content_html']
3165             more_widget_html = more['load_more_widget_html']
3166
3167     def _real_extract(self, url):
3168         page = self._download_webpage(
3169             'https://www.youtube.com/feed/%s' % self._FEED_NAME,
3170             self._PLAYLIST_TITLE)
3171         return self.playlist_result(
3172             self._entries(page), playlist_title=self._PLAYLIST_TITLE)
3173
3174
3175 class YoutubeWatchLaterIE(YoutubePlaylistIE):
3176     IE_NAME = 'youtube:watchlater'
3177     IE_DESC = 'Youtube watch later list, ":ytwatchlater" for short (requires authentication)'
3178     _VALID_URL = r'https?://(?:www\.)?youtube\.com/(?:feed/watch_later|(?:playlist|watch)\?(?:.+&)?list=WL)|:ytwatchlater'
3179
3180     _TESTS = [{
3181         'url': 'https://www.youtube.com/playlist?list=WL',
3182         'only_matching': True,
3183     }, {
3184         'url': 'https://www.youtube.com/watch?v=bCNU9TrbiRk&index=1&list=WL',
3185         'only_matching': True,
3186     }]
3187
3188     def _real_extract(self, url):
3189         _, video = self._check_download_just_video(url, 'WL')
3190         if video:
3191             return video
3192         _, playlist = self._extract_playlist('WL')
3193         return playlist
3194
3195
3196 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
3197     IE_NAME = 'youtube:favorites'
3198     IE_DESC = 'YouTube.com favourite videos, ":ytfav" for short (requires authentication)'
3199     _VALID_URL = r'https?://(?:www\.)?youtube\.com/my_favorites|:ytfav(?:ou?rites)?'
3200     _LOGIN_REQUIRED = True
3201
3202     def _real_extract(self, url):
3203         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
3204         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, 'favourites playlist id')
3205         return self.url_result(playlist_id, 'YoutubePlaylist')
3206
3207
3208 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
3209     IE_DESC = 'YouTube.com recommended videos, ":ytrec" for short (requires authentication)'
3210     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/recommended|:ytrec(?:ommended)?'
3211     _FEED_NAME = 'recommended'
3212     _PLAYLIST_TITLE = 'Youtube Recommended videos'
3213
3214
3215 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
3216     IE_DESC = 'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
3217     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
3218     _FEED_NAME = 'subscriptions'
3219     _PLAYLIST_TITLE = 'Youtube Subscriptions'
3220
3221
3222 class YoutubeHistoryIE(YoutubeFeedsInfoExtractor):
3223     IE_DESC = 'Youtube watch history, ":ythistory" for short (requires authentication)'
3224     _VALID_URL = r'https?://(?:www\.)?youtube\.com/feed/history|:ythistory'
3225     _FEED_NAME = 'history'
3226     _PLAYLIST_TITLE = 'Youtube History'
3227
3228
3229 class YoutubeTruncatedURLIE(InfoExtractor):
3230     IE_NAME = 'youtube:truncated_url'
3231     IE_DESC = False  # Do not list
3232     _VALID_URL = r'''(?x)
3233         (?:https?://)?
3234         (?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/
3235         (?:watch\?(?:
3236             feature=[a-z_]+|
3237             annotation_id=annotation_[^&]+|
3238             x-yt-cl=[0-9]+|
3239             hl=[^&]*|
3240             t=[0-9]+
3241         )?
3242         |
3243             attribution_link\?a=[^&]+
3244         )
3245         $
3246     '''
3247
3248     _TESTS = [{
3249         'url': 'https://www.youtube.com/watch?annotation_id=annotation_3951667041',
3250         'only_matching': True,
3251     }, {
3252         'url': 'https://www.youtube.com/watch?',
3253         'only_matching': True,
3254     }, {
3255         'url': 'https://www.youtube.com/watch?x-yt-cl=84503534',
3256         'only_matching': True,
3257     }, {
3258         'url': 'https://www.youtube.com/watch?feature=foo',
3259         'only_matching': True,
3260     }, {
3261         'url': 'https://www.youtube.com/watch?hl=en-GB',
3262         'only_matching': True,
3263     }, {
3264         'url': 'https://www.youtube.com/watch?t=2372',
3265         'only_matching': True,
3266     }]
3267
3268     def _real_extract(self, url):
3269         raise ExtractorError(
3270             'Did you forget to quote the URL? Remember that & is a meta '
3271             'character in most shells, so you want to put the URL in quotes, '
3272             'like  youtube-dl '
3273             '"https://www.youtube.com/watch?feature=foo&v=BaW_jenozKc" '
3274             ' or simply  youtube-dl BaW_jenozKc  .',
3275             expected=True)
3276
3277
3278 class YoutubeTruncatedIDIE(InfoExtractor):
3279     IE_NAME = 'youtube:truncated_id'
3280     IE_DESC = False  # Do not list
3281     _VALID_URL = r'https?://(?:www\.)?youtube\.com/watch\?v=(?P<id>[0-9A-Za-z_-]{1,10})$'
3282
3283     _TESTS = [{
3284         'url': 'https://www.youtube.com/watch?v=N_708QY7Ob',
3285         'only_matching': True,
3286     }]
3287
3288     def _real_extract(self, url):
3289         video_id = self._match_id(url)
3290         raise ExtractorError(
3291             'Incomplete YouTube ID %s. URL %s looks truncated.' % (video_id, url),
3292             expected=True)